PDA

View Full Version : How - In Just 72 Hours - I Used PPC To Stop My Client From Going Broke


Colin Parker
23rd January 2010, 08:05
Here is an example of the power of Google Adwords - used correctly.

I have just started a JV with a new client/potential business partner - he offers a great service in a very competitive market.

His problem is he needs more leads - and fast - because his competitors are increasingly using SEO and PPC to gain market share and his sales volume is dropping to near crisis point.

Leads for his service are sold on the internet for £45-£65 per lead. At this cost he cannot afford to buy the volume leads he needs to make the numbers crunch in his business.

My agreement with him was this: he pays for the cost of Google Adwords - he gets leads at my cost price plus a relatively small monthly management fee - and he pays me a commission on resulting sales.

He agreed and I started split testing the Google PPC advertising campaign on Tuesday of this week.

Before we 'floor the pedal' on Google Adwords advertising spend I wanted to test all aspects of the campaign: cost per click - ad content - ad position - website conversion. By doing this I can quickly establish a cost per lead and we can then make a decision on how many leads we want each week/month and how much that will cost.

In just three days - from Wednesday to close Friday - I was able to do a collective total of 73 different tests on the campaign. Google's Website Optimiser, Conversion Statistics and Analytics Tool allows me to constantly measure the results of these tests - and incrementally step by step increase the campaign conversion and reduce the lead cost.

The results of these split tests - in just 3 days - were as follows:

My 'click thru rate' (the % of potential clients who click on my ad as opposed to my competitors) increased 7 fold.

My 'cost per click' (the price I pay when someone clicks on my ad) dropped by 36%

My 'average ad position' (there are 11 ads on the page - the higher the position generally the more clicks) increased from 6.6 to 1.8. 1.8 puts me in the coveted top 3 positions directly above the SEO ads on the left of the page - and by so doing one of our biggest competitors has been displaced.

My 'website conversion' (the number of people who submitted their contact details on the site) increased from 11.8% to 56.4%. A near five fold increase.

My 'cost per lead' (click cost divided by number of contact submissions) dropped from £23.55 to £5.72. A near 75% drop.

All of this in just 3 days.

At £5.72 per lead my client/partner is now getting between 8 and 12 leads for the price he would have paid for 1 from lead sellers.

I also believe that by the end of next week further split testing will reduce the cost per lead to below £4.

We have now decided to generate 200 leads a week - advertising spend c£1,000 compared to c£9,000-£13,000 if purchased from leads sellers.

I expect to earn a six figure annual commission from my part of the deal - and my client/partners business will be transformed. To a degree it already has been. He is certainly sleeping easier now that's for sure.

That is the power of Google Adwords - fast, measurable and capable of taking you from 'zero to hero' in just 72 hours.

Colin Parker

Tej
23rd January 2010, 08:35
Wow.. fantastic.. Zero to Hero in 72 hours.

I am totally naive about websites, SEO, adwords, etc....

But am interested to know ... does leads =actual business? ( income)

The spend might have been reduced to £1000... but its still spend.

Has the income increased dramatically for your client and yourself ( expected income 6 figures etc) to sleep easily... in 72 hours!! ( you get a management fee anyway.. so guess you sleep slightly better)

Really powerful stuff... seems adwords could be the saving of all businesses that are in sh1t creek... and those that are doing well, to be doing miles better.

Why don't all business owners follow this simple solution?

just a thought..

Webtistic
23rd January 2010, 08:50
A consistent 56.4% conversion rate for a highly competitive sector (I am guessing finance/insurance) is somewhat amazing.

How about letting us know what the landing page is so we can see what a perfect conversion set-up looks like? Maybe your client will even get a few more leads! :)

dots and spots Jeff
23rd January 2010, 09:06
Some really great stats and I have no doubt that you have done a great job.

To get the complete picture it'd be good to know how many of those 200 leads per week end up buying, and how much they spend - i.e. how much cash does his £1000 a week spend generate.

seedstotal
23rd January 2010, 09:12
near five fold increase in conversion dots and spots, so i guess they cant complain, he is not gonna tell you how much he earns, come on.
Great seminar Coli, i guess :), i wish i could go, really could do with a five fold increase :)

Colin Parker
23rd January 2010, 09:59
A consistent 56.4% conversion rate for a highly competitive sector (I am guessing finance/insurance) is somewhat amazing.

How about letting us know what the landing page is so we can see what a perfect conversion set-up looks like? Maybe your client will even get a few more leads! :)

Dave - the last thing I am going to tell ANYONE is what the landing page is!

That is like giving a burglar the key to your safe.

One of the biggest disadvantages of IM is that if a competitor knows your website is a high converter it can be replicated.

I would NEVER therefore give away specific figures for a specific campaign.

It is akin to marketing suicide.

Colin Parker

dots and spots Jeff
23rd January 2010, 10:34
near five fold increase in conversion dots and spots, so i guess they cant complain, he is not gonna tell you how much he earns, come on.


Fair enough - but he was happy enough to tell us how much he spends.

I know its hard to discern motives in the written media of a forum, but I was just genuinely intersted in this as a story - spending £50k a year on advertising is, I'm afraid, way out of our league!

Like Tej, I too am

I am totally naive about websites, SEO, adwords, etc...
and was just curious to see what an advertising spend of £1000 a week generates - we could all go out and spend that it whatever form we wanted to, the question is whether I could use it to generate a profit! :)

As the OP says, the aim is to get people to submit their contact details and Colin, the OP has clearly done a great job here and got the cost per contact down to about £5 from about £50, so he has met his brief and done a great job.

But the business he is working for still has to convert those leads into sales. Drilling down into the figures a little more, lets say he pays £1000 a week for 200 leads. If those leads then convert to buyers at a rate of 5%, then that's 10 buyers have cost him £1000, so each buyer needs to spend £100 just to cover the cost of the PPC campaign.

Now it may well be that the actual figures are much better than those quoted above, I don't know (I hope they are) it would have just added a little more flesh to a great story to have that detail, but, as I said above, well done Colin - I think you have shown what a good (professionally run) adwords campaign can do.

(Just another quick thought before I press send - if the business that commissioned the OP can't turn the leads into a profit, he could always sell those leads on to his competitors:


We have now decided to generate 200 leads a week - advertising spend c£1,000 compared to c£9,000-£13,000 if purchased from leads sellers.
and turn that £1000 a week into £9000!)

Tej
23rd January 2010, 10:45
IMO

Anyone can spout figures... means nothing... could be a load of BS.

I do read a lot of posts.. and the person who I consider to be very good.... Steve Gibson... always backs up what he says.

Provides source etc...

The rest of this stuff is mickey mouse IMO

Colin Parker
23rd January 2010, 11:24
IMO

Anyone can spout figures... means nothing... could be a load of BS.

I do read a lot of posts.. and the person who I consider to be very good.... Steve Gibson... always backs up what he says.

Provides source etc...

The rest of this stuff is mickey mouse IMO

Sorry ... but you are dead wrong about Steve Gibson.

Steve is a total professional and exceptionally good at what he does.

I know because I have had a number of conversations with him and we both know enough about IM - and Google Adwords and website conversion especially - to know that neither of us talks BS.

So ... why are you dead wrong about Steve?

Because - like me - he would NEVER reveal the name or website of a particular client when he gives conversion data etc., in his posts.

That would be a) a breach of client confidentiality and b) marketing suicide as I have said previously.

With respect - I couldn't care less what you believe about the figures I have put up.

I am trying to motivate forum members to learn more about PPC and how it can spectacularly increase their business. And Steve Gibson would be a great guy to go to for help.

The figures are 100% factual. If you don't believe them - that is your choice. Because I aint ever going to reveal the client and website.

Colin Parker

Tej
23rd January 2010, 11:37
I have not asked about revealing your site.. means nothing to me

My business sense tells me

If you have "decided to generate 200 leads".. and this costs £1000.
You can then sell these leads at £50 a pop=£10k
nett profit £9000.

If I were the client..what other biusiness would I need... I would happily pay you £5000, and plus even add your "small management fee" of £1000 per week

I would be more than happy to have £3000 every week in my pocket

Sleep well... happyy days!!

Does that scenario work? ( rough figures obviously)

SteveGibson
23rd January 2010, 11:38
really could do with a five fold increase

You don't need a five fold increase to make this worthwhile. Even aiming for 10% here and there and you'll see what a difference it makes.

I made a post a few days ago (http://www.ukbusinessforums.co.uk/forums/showthread.php?t=140467) where I showed how improving 5 things about adwords by just 10% will increase your adwords income by 233%.

As Colin has shown from his results, 10% isn't a big deal. 30% or 100%+ can be quite possible.

(anyone who has split-test Adwords ads will know this - it's fairly common to go from a 1% ctr to 1.5% or 2.0%)

So, if you're thinking "I don't know how to increase my conversion rate by 400%", just set up a split-test and try to increase it by 10%.

Lao-tzu said "a journey of a thousand miles begins with a single step".

Woody Allen said "80 percent of success is just showing up".

Most testing success isn't about doubling or tripling a metric, it's about testing relentlessly - i.e. "just showing up" and testing - and having those 10%-30% improvements multiply like compound interest.

So don't let Colin's story intimidate you into thinking testing is about being brilliant.

It's usually about knowing you're not brilliant, you're not all-knowing, and asking your prospects (via a test) what they'll respond to.

That's why I test. That's why Colin tests.

We could both guess. And we've both got the experience where we could guess pretty well.

But we make so much more money by letting our prospects tell us what they want - in a PPC ad, on a web page - than if we try to dictate to them.

Steve

SteveGibson
23rd January 2010, 11:50
Because - like me - he would NEVER reveal the name or website of a particular client when he gives conversion data etc., in his posts.

That's right.

The only time I've ever shown the before & after of split-tests is my own sites and emails.

(Like my thread a couple of weeks ago where I showed 3 split-tests of email subject lines)

I don't give the explicit details of client stuff. There are some testimonials on my sites where you can read client comments about increases in conversion rate. And there are some screen-shots of split-test results from Website Optimizer.

But the details of the changes are confidential.

Competitors copy each other.

Just this week, I was using archive.org, spyfu and alexa to try to work out why a client's key competitor was able to be top of the PPC listings. What did they do to their site that allowed them to step up and take over?

I think I figured it out.

My client made the same change to his site and, within 2 hours of that change going live, he was raving about the increase in enquiries.

If I can do that without being able to see the competitor's test data, what do you think I could do if they made it publicly available?

Steve

Scott-CopyandDesign
23rd January 2010, 11:53
Great thread. Stuff like this reminds me that I need to get a DM venture under way. Just need to find the time to do it.

saxondale
23rd January 2010, 11:57
My client made the same change to his site and, within 2 hours of that change going live, he was raving about the increase in enquiries.


Steve


I`m raving about the changes I`ve made ................ my enquiries are up 300%















raving mad, for some reason google now thinks I`m the imigration agency and I`m getting all their calls





.

Colin Parker
23rd January 2010, 11:58
I have not asked about revealing your site.. means nothing to me

My business sense tells me

If you have "decided to generate 200 leads".. and this costs £1000.
You can then sell these leads at £50 a pop=£10k
nett profit £9000.

If I were the client..what other biusiness would I need... I would happily pay you £5000, and plus even add your "small management fee" of £1000 per week

I would be more than happy to have £3000 every week in my pocket

Sleep well... happyy days!!

Does that scenario work? ( rough figures obviously)

You are assuming that:

a) my client wants to operate a lead selling business - he doesn't.
b) even if he did that I would not cut him out of the deal and go direct.
c) the increased profit from selling the leads within his own business would even make selling the leads an option.
d) selling leads is a consistent way to make money - it isn't - lead buyers are notoriously fickle and often don't pay you if they don't sell.
e) he might be as sceptical and have the same loser mentality that you do - he doesn't.

Colin Parker

saxondale
23rd January 2010, 12:00
You are assuming that:

a) my client wants to operate a lead selling business - he doesn't.
b) even if he did that I would not cut him out of the deal and go direct.
c) the increased profit from selling the leads within his own business would even make selling the leads an option.
d) selling leads is a consistent way to make money - it isn't - lead buyers are notoriously fickle and often don't pay you if they don't sell.
e) he might be as sceptical and have the same loser mentality that you do - he doesn't.

Colin Parker


I made it more noticable incase any future clients miss it -



commercial suicide in real time .................priceless

cmcp
23rd January 2010, 12:09
Good inspiration, thanks for sharing.

Tej
23rd January 2010, 12:11
You are assuming that:

a) my client wants to operate a lead selling business - he doesn't.
b) even if he did that I would not cut him out of the deal and go direct.
c) the increased profit from selling the leads within his own business would even make selling the leads an option.
d) selling leads is a consistent way to make money - it isn't - lead buyers are notoriously fickle and often don't pay you if they don't sell.
e) he might be as sceptical and have the same loser mentality that you do - he doesn't.

Colin Parker



Sorry ... I have assumed nothing.... I asked a simple question.. and gave you a simple scenario....without any technical crap trap ( thats chinese for clap trap)...


You are the one asuming....

If £3000 every week is a loser mentality... happy days!!

( I am wishing that you do double that regularly)

... surprised that you have time/inclination to post on this forum with that kinda money every week in your bin

:D:D

adventurelife
23rd January 2010, 12:18
PPC is by far the best marketing tool I have every seen. It provides instant research and results both positive and negative in super quick time which no other route to market does.

I used it to research and launch my business years ago. I did it very , very badly and burned through over £12k per year for first few years. Even doing it very badly it still paid got the business going and paid the wages and gave me breathing space to get my other marketing up to speed.

Those that can do it correctly I suspect can bring great value to their business. I also suspect to be really good at it now with all the competition it requires huge amount of time using and testing and hence why I do not use as I would be a rank amateur competing with accounts managed by those that know what they are doing.

200 leads from one of our target sectors would make me a very happy man:) as we convert 55% of our leads at a revenue level current and 66% at a booking level.

Colin Parker
23rd January 2010, 13:02
... surprised that you have time/inclination to post on this forum with that kinda money every week in your bin

:D:D

So what is your point ... only failures post on this forum? Is that it?

I often do wonder why I post on this forum - and attitudes like yours are one of the reasons why.

Read my previous threads and posts - check the number of thanks I have had ...

I am either the biggest BSer ever or I tell it as it is.

I'll leave others to make their minds up - but as I have said it really does not matter at all what you or anyone else believes.

Myself and my clients know the truth ...

And that is ALL that really matters to me.

Your opinions are totally insignificant.

Colin Parker

Colin Parker

Tej
23rd January 2010, 13:16
Why are you getting all worked up?

I have asked a couple of simple questions.. and you have had difficulty in answering them.

Read my initial post.. you have ignored it completely!!

You take one sentence out of context.. and spout further BS

Either give me an answer to a simple question or be a man, and say you have no answer..

its pretty simple isn't it?

You abuse me with all this loser mentality crap..

you can't take simple can you?

I have not disputed anything you have said... i dont do technical....

papverpoppies
23rd January 2010, 13:29
I don't do techincal either, but if it is that easy - and more to the point that fast -why does everyone who has the know how, not do it for their clients?

Are you saying, that you can turn anyones business around in 72 hours or whatever, it all sounds too good to be true, so why are people not queueing up for this service.

Only asking in these terms, as I do not know how else to put it across, as a non techi!:|

Poppy

Tej
23rd January 2010, 13:35
I don't do techincal either, but if it is that easy - and more to the point that fast -why does everyone who has the know how, not do it for their clients?

Are you saying, that you can turn anyones business around in 72 hours or whatever, it all sounds too good to be true, so why are people not queueing up for this service.

Only asking in these terms, as I do not know how else to put it across, as a non techi!:|

Poppy

Poppy

Thats similar to what I said in my very first post:)

He does not have any answers to simple questions.... looks like:)


and then will come the age old stuff

" I often do wonder why I post on this forum - and attitudes like yours are one of the reasons why."

:D:p:D

Scott-CopyandDesign
23rd January 2010, 13:55
Really powerful stuff... seems adwords could be the saving of all businesses that are in sh1t creek... and those that are doing well, to be doing miles better.

Why don't all business owners follow this simple solution?

I don't do techincal either, but if it is that easy - and more to the point that fast -why does everyone who has the know how, not do it for their clients?

Are you saying, that you can turn anyones business around in 72 hours or whatever, it all sounds too good to be true, so why are people not queueing up for this service.

Only asking in these terms, as I do not know how else to put it across, as a non techi!:|

Poppy

Because they don't realise how worthwhile it is. They're stuck in their ways and only see value on spending their budget on advertising. Unfortunately, it's common for the resulting traffic to convert poorly.

It's natural human behaviour causing the 'too good to be true' scepticism which is a shame. I often have to explain how worthwhile conversion optimisation is (double the conversion, double the sales without any extra traffic). Mind you, these days I'm noticing more and more business owners contacting me to look at optimisation. Seems to thankfully be catching on.

Business owners CAN follow the simple solution. They can either spend time learning how to do it themselves, or pay a specialist to do it for them. People who have the know how do actually do it for their clients. I improve conversion rates all the time. Mind you, it's rare to improve the rate to such a degree as Colin has done. It does depend on the state of the website prior to optimisation.

Many people just don't take the incentive to get it looked into.

papverpoppies
23rd January 2010, 13:55
Poppy

Thats similar to what I said in my very first post:)

He does not have any answers to simple questions.... looks like:)


and then will come the age old stuff

" I often do wonder why I post on this forum - and attitudes like yours are one of the reasons why."

:D:p:D

I was watching TV very very late last night, the part where they run the
gimmick ads - that will make us all, super thin, rich, beautiful, etc etc, and all for the amazing price of ££££ plus p/p.

Then they get the super thin 'already' beautiful models to tell you how this 'product' worked for them (like they ever used it in their lives)!

Or the business manual that will make you wonder how you ever 'survived' with out it!

My fav is that product that can make you lay a course of bricks in a staight line (just like the pro bricky)...and they show some women in her 60s slapping down the bricks!
What they dont say is you have got to mix the mortar, cut the bricks, you get the picture.

If everything was that easy - would we all not be beautiful, thin, building our own homes and running highly successful business's?

If it IS that easy - lead me to it - I have no probs with laying a few bricks.;)

Poppy

papverpoppies
23rd January 2010, 13:59
Because they don't realise how worthwhile it is. They're stuck in their ways and only see value on spending their budget on advertising. Unfortunately, it's common for the resulting traffic to convert poorly.

It's natural human behaviour causing the 'too good to be true' scepticism which is a shame. I often have to explain how worthwhile conversion optimisation is (double the conversion, double the sales without any extra traffic). Mind you, these days I'm noticing more and more business owners contacting me to look at optimisation. Seems to thankfully be catching on.

Business owners CAN follow the simple solution. They can either spend time learning how to do it themselves, or pay a specialist to do it for them. People who have the know how do actually do it for their clients. I improve conversion rates all the time. Mind you, it's rare to improve the rate to such a degree as Colin has done. It does depend on the state of the website prior to optimisation.

Many people just don't take the incentive to get it looked into.


So does it also depend on the ability of the person to be able to make this happen?

If I did not find a Colin type person who was capable and understood this process, would I, could I, be loosing out.

How do you (as a non techi) person make sure you get the right person?

Poppy

Tej
23rd January 2010, 14:10
. Mind you, it's rare to improve the rate to such a degree as Colin has done.



Thank you Scott... I now know a bit more:)

I have taken a sentence of your quote out of context, like the OP did to mine...

what does that make him out to be?

:D:D

SteveGibson
23rd January 2010, 14:17
I don't do techincal either, but if it is that easy - and more to the point that fast -why does everyone who has the know how, not do it for their clients?

Speaking as someone who tries to get his clients to split-test, it's usually like banging your head against a brick wall. I've got a couple of clients that'll test, but lots more that should test, but just don't want to.

So let me address a bigger question: why don't people split test?

I've been thinking about this for a while and here are the reasons I've come up with:

(1) Seems "too easy" - we're brought up to think "hard work= money", so when someone changes one sentence on their website and gets a 20%-30% pay rise, that doesn't fit our model of how to make money.

(2) People don't trust numbers - it's probably no coincidence that Colin and I have both worked in financial services. We're both from a background where "the numbers never lie". So we believe what we see.

Most people aren't that comfortable with numbers and the sorts of stuff I posted (the 233% increase from 5 improvements of 10%) makes them uncomfortable. "How can 5 x 10% = +233%?"

(3) Social proof - when we're unsure, we humans tend to look to people who are similar to ourselves and do what they do. On this forum we have dozens of people going on and on about SEO. We have 2 people talking about split-testing.

It's normal to think, "If split-testing was so good, more people would be doing it. More people aren't doing it, therefore it must be bollox."

(4) They think it's difficult - there's an expectation that the technical side of setting up a test is going to be confusing. In reality, Google Website Optimizer is easier to understand than Google Analytics (and not much harder to set up).

(5) They're being held back by web designers - many business owners are dependent on their web designers to do the techie stuff for them. However, most web designers have no understanding of how to do a split-test. So, the designers are reluctant to help.

In addition, some of them don't want their site to be split-tested, either because it's their "baby" or because they feel they'll be undermined when some of their design decisions are shown to be costing the client a bucketful of money.

(6) They just don't understand why it's necessary - the common things I hear from clients are "why do we have to test it, can't we just change it and see what happens?" or "are you sure it's going to work?".

They don't get that sequential testing (running one version for a while then running the other version) produces unreliable results. (particularly when the differences are less than 20%)

Nor do they seem to get that testing isn't about guaranteed winners. If you're winning all your tests, you must have a crap site (one with obvious weaknesses). Testing is about getting the answers to questions that can go either way.

Those 6 reasons are my half-arsed speculation about why people don't test. Or, to put it another way, those are the most logical reasons I can think of to explain why people behave illogically.

Steve

Colin Parker
23rd January 2010, 14:18
Then they get the super thin 'already' beautiful models to tell you how this 'product' worked for them (like they ever used it in their lives)!

If everything was that easy - would we all not be beautiful, thin, building our own homes and running highly successful business's?

If it IS that easy - lead me to it - I have no probs with laying a few bricks.;)

Poppy

Thing is Poppy ... there isn't any easy way to lose weight.

As a lot of fat lasses and lads will tell you - if you can understand what they are saying while chewing a double cheesburger.

It takes months (normally) of proper dieting and perhaps exercise to get anywhere near the shape of those models.

And there isn't any easy way to get the PPC results I do.

Never mind months - I have spent nearly 30 years in direct marketing and sales and 80% of what I learnt and applied offline also applies online.

The other 20% for online I have learnt over the past 10 years - spending literally 100's of £1,000's on PPC (a big 6 figure total) - buying loads of ecourses (mostly rubbish) - and meeting and learning from some of America's most successful online marketeers.

So ... don't believe anything is easy. For a long time it wasn't for me.

But after nearly 30 years in direct marketing - and with 10 years of IM experience behind me - I'm more than starting to get the hang of it.

As my original post on this thread shows.

Colin Parker

Scott-CopyandDesign
23rd January 2010, 14:18
So does it also depend on the ability of the person to be able to make this happen?

If I did not find a Colin type person who was capable and understood this process, would I, could I, be loosing out.

How do you (as a non techi) person make sure you get the right person?
Poppy

The same way you make sure you get anything else of quality. Look at the figures and look at the testimonials. Make the effort to understand the technical side so you can choose the right person for the job. You will only ever be rewarded for doing so.

Mind you, anyone selling their services should always explain everything in easily absorbent ways. Colin is only showing an example of what he did and how he did it. It's valuable information.

Thank you Scott... I now know a bit more:)

I have taken a sentence of your quote out of context, like the OP did to mine...

what does that make him out to be?

:D:D

It still stands out of context. Colin is only showing an example of Adwords and the potential power for any business. He isn't selling his services and claiming he can get this kind of improvement for everyone out there. Is a five fold increase in conversion possible? Yes.

Tej
23rd January 2010, 14:27
Steve.. a simple question.... I don't understand it to be honest...

main points I would make.

Do people actually know and understand what " split testing" is? ( I know I don't)

and would they be able to do it themselves if they had the inclination to do so?

personally... if I had someone who did the marketing for me.. did any campaigns for me etc.. I would rely on that person 100%... give him the responsibility and authority.. to achieve results. A kind of a carte blanche .. the limitation being the budget.

Does that make sense?

Does that make me a lazy barsteward?

:D

Tej
23rd January 2010, 14:32
It still stands out of context. Colin is only showing an example of Adwords and the potential power for any business. He isn't selling his services and claiming he can get this kind of improvement for everyone out there. Is a five fold increase in conversion possible? Yes.

That wasn't what I asked... but no matter.. I understand:)

papverpoppies
23rd January 2010, 14:34
Steve.. a simple question.... I don't understand it to be honest...

main points I would make.

Do people actually know and understand what " split testing" is? ( I know I don't)

and would they be able to do it themselves if they had the inclination to do so?

personally... if I had someone who did the marketing for me.. did any campaigns for me etc.. I would rely on that person 100%... give him the responsibility and authority.. to achieve results. A kind of a carte blanche .. the limitation being the budget.

Does that make sense?

Does that make me a lazy barsteward?

:D

That is the angle I would be coming in from.

What if I did not get a Colin, Steve or Scott type person, how would I, the non techinical person, know I could be achieving better results.

That is the scary part.

Poppy

dots and spots Jeff
23rd January 2010, 14:43
I have been following this thread (and other similar ones with interest) and I have come to the conclusion that little changes can have a big impact, but its not always easy to know, in advance, what change will have which impact and so you split test to discover this. I understand a split test to be something like this:

You have 2 pages that are identical, with one small change: e.g.

Page A may say "click here if you love chocolate"
Page B may "Chocolate Lover? Click Here"

and count what percentage of visitors click the link on Page A and compare to what percentage of vistiors click the link on Page B

I like numbers, and with a mathematical/scientific/engineering background the principles behind this seem pretty straight forward (although I appreciate the practice may be more complicated!) and, because they allow you to measure things, take away a lot of guess work, helping to do away with the old adage "I know half of my advertising budget is wasted, but I don't know which half". (I write the above not as an expert, but as someone who is trying to summarise their own understanding - I won't be offended if anyone says I've got it all wrong!)

I'd like to learn more about split testing, conversion testing etc. Can anyone recommend any good books, courses, projects I could undertake to help me find out more. Presumably you need a minimum sample size for the testing to be valid - is there a ball park figure? Any advice/pointers/help would be most welcome!

Thanks to all for a really interesting and thought provoking thread!

Jeff

Mystro
23rd January 2010, 14:45
Thats a great example of a the right marketing for your business, marketing is not a one shoe fits all strategy.

Many things in marketing don't makes sense of why it works for one person and not for another, split testing is somthing i dont unserstand either, i get the concept of what it is and why its important when trying to create the right marketing strategy as its so easy to loose money on somthing just because you think it will work for you because Joe Blogs earned a fortune doing it ..

There seems to be a lot of post where people take the time to share some of their better learning experiences only to get bombarded with critisism or doubt.

Not pointing the finger at anyone in particular, just making an observation.

Tej
23rd January 2010, 14:46
That is the angle I would be coming in from.

What if I did not get a Colin, Steve or Scott type person, how would I, the non techinical person, know I could be achieving better results.

That is the scary part.

Poppy

LOl.. its all surrounded in mystique... cloak and dagger kinda stuff...

I guess its geared that way... every discipline to have its own peculiarity.

Else people like the OP who has spent hundreds of thousands learning the art, would not be able to ply his trade and stop his clients from going broke in just 72 hours.

Gobsmacking time.. don't you think?

:)

SteveGibson
23rd January 2010, 14:46
Steve.. a simple question.... I don't understand it to be honest...

main points I would make.

Do people actually know and understand what " split testing" is? ( I know I don't)

If you've used adwords and split-test adwords ads, then you know what split-testing is.

Doing it with a web page is the same concept.

and would they be able to do it themselves if they had the inclination to do so?

Search for info on "Google Website Optimizer". Google have plenty of info on their site. They have videos on how to use it on youtube.

It's not a difficult piece of software to use.

personally... if I had someone who did the marketing for me.. did any campaigns for me etc.. I would rely on that person 100%... give him the responsibility and authority.. to achieve results. A kind of a carte blanche .. the limitation being the budget.

Does that make sense?

Yes.

You could learn how to do it yourself - and it's really not that hard, so I'd encourage people to find out about it - but hiring a marketer who knows about split-testing has its advantages:

(1) They've done a lot of testing and worked with a number of sites. This means they've developed an idea of "best practice" - i.e. the things that work more often than not. That experience can identify the best elements (on the page) to test and the best alternatives to test against them.

(2) If they have copywriting skill, they can test their copy against the existing copy and will beat it more often than not. The better their copy, the bigger the gains.

(3) They'll help with the techie side of using website optimizer. The first time I set up a test, it didn't work. (it's ok, it was on one of my own sites - I made sure I was my own guinea pig) I soon realised what I'd done. But there are other things I've figured out so my clients don't have to.

(there are some very old threads in the private section where I asked people to visit urls and tell me what they saw - that was to do with split-testing over multiple pages)

(4) If the marketer has a stats background, then he'll know how to set complex tests up so you're testing the right things. (when you're testing more than one element at a time)

(5) If the marketer understands testing, he'll know how to interpret the results and when to call a result. (to do with statistical confidence)

(6) It's just good to have an outside pair of eyes. Some of my test ideas have come from looking at a site and not understanding something. So, we test a version of the page with an explanation, or with clearer buying instructions.

I don't want this to be a sales pitch. I'll say it again: it's easy enough to learn about this. And, ideally, you should watch a couple of videos even if you're thinking about outsourcing it. There are some where google show their own testing and results.

And, because it's google talking about their own pages, I'm sure it has far more credibility than a couple of guys on a forum.

Steve

SteveGibson
23rd January 2010, 14:54
I have been following this thread (and other similar ones with interest) and I have come to the conclusion that little changes can have a big impact, but its not always easy to know, in advance, what change will have which impact and so you split test to discover this. I like numbers, and with a mathematical/scientific/engineering background the principles behind this seem pretty straight forward (although I appreciate the practice may be more complicated!) and, because they allow you to measure things, take away a lot of guess work, helping to do away with the old adage "I know half of my advertising budget is wasted, but I don't know which half". (I write the above not as an expert, but as someone who is trying to summarise their own understanding - I won't be offended if anyone says I've got it all wrong!)

You're kinda right.

One way to think of a test is as a "forced survey". The problem with traditional surveys is that people give answers that are unreliable - i.e. what they say they'd do, or think they'd do, isn't what they'd actually do.

With a test, you're really asking the question "would you buy/opt-in if you saw this version of the page?" and, because the user doesn't know he's being surveyed, he behaves normally.

And you're doing that with two pages (original and test version) and then comparing the conversion rates.

I'd like to learn more about split testing, conversion testing etc. Can anyone recommend any good books, courses, projects I could undertake to help me find out more.

Go to youtube and search on "website optimizer". Watch the videos by google themselves.

Presumably you need a minimum sample size for the testing to be valid - is there a ball park figure? Any advice/pointers/help would be most welcome!

Website optimizer tells you the percentage chance each version has of winning. I usually work to 90% statistical confidence.

Here's a link to screenshot of a test I did:

http://www.stevegibsonconsulting.co.uk/optimiser.jpg

you'll see the info Website Optimizer gives.

Steve

papverpoppies
23rd January 2010, 15:46
Thats a great example of a the right marketing for your business, marketing is not a one shoe fits all strategy.

Many things in marketing don't makes sense of why it works for one person and not for another, split testing is somthing i dont unserstand either, i get the concept of what it is and why its important when trying to create the right marketing strategy as its so easy to loose money on somthing just because you think it will work for you because Joe Blogs earned a fortune doing it ..

There seems to be a lot of post where people take the time to share some of their better learning experiences only to get bombarded with critisism or doubt.

Not pointing the finger at anyone in particular, just making an observation.

If you do not understand something - then you have to ask - surely.

The human race thrives on critisism and doubt, no human got injured in the making of this thread!;)

I have learnt a 'little' not because of the inability of the posters to this thread, just my lack of knowledge on this matter.

It is an interesting subject matter.

Thanks

Poppy

AaronB
23rd January 2010, 17:39
I must agree with a lot of people "your" first post is like Cardell Media ( chris cardell inner circle) email I get, a blog of how good his marketing is,.

Aaron

Colin Parker
23rd January 2010, 18:44
I must agree with a lot of people "your" first post is like Cardell Media ( chris cardell inner circle) email I get, a blog of how good his marketing is,.

Aaron

Aaron how many of my other 400 posts have you read?

The ones that have got me nearly 200 thanks?

And as for Chris Cardell and myself showing 'how good our marketing is' - what do you us to say - how bad it is?

If I'm about to have an operation I don't want the surgeon to tell me he is an inexperienced amateur who will 'give it a go' ...

The original post was an example of how split testing on Google Adwords can have a massive and virtually immediate impact on a business.

It was put up to motivate people to educate themselves and learn about PPC to help them achieve what I imagine most want - more sales and profits and more money in the bank.

However - as always - the doubters, sceptics and losers can't wait to weigh in with abject negativity. And in so doing they demotivate those who - perhaps like you - are open minded to change and learning.

Colin Parker

Colin Parker
23rd January 2010, 19:24
LOl.. its all surrounded in mystique... cloak and dagger kinda stuff...

I guess its geared that way... every discipline to have its own peculiarity.

Else people like the OP who has spent hundreds of thousands learning the art, would not be able to ply his trade and stop his clients from going broke in just 72 hours.

Gobsmacking time.. don't you think?

:)


Lets get this into context ...

I did not literally stop my client going broke in 72 hours.

Had he continued with the same lead generation model his business would have continued to decline and HE was worried that could have lead to him going out of business.

In 72 hours I gave him the leads that now make going out of business far less likely - in that sense I have stopped him going broke.

However - the HEADLINE written for this thread 'How - In Just 72 Hours - I Used PPC To Stop My Client From Going Broke' is actually proof positive of the power of split testing.

This thread had nearly 700 views by 6pm today (Saturday).

It is the most viewed thread on the PPC/SEO page today with the exception of Steve Gibson's 'Five Deadly Sins' thread which has been up for four days.

It has therefore had more views than any other thread in this section today - BY FAR.

Had I put up a thread headline like 'A Split Testing Example' I bet the views would not have been more than 50-100.

That is 7-14 times less views than this thread has had.

I put the thread headline up because I know - from testing 1.000's of headlines over the years - which headlines get attention and which ones don't.

And EXACTLY the same applies when you write a PPC ad headline.

That is how I increased my CTR seven fold in just 72 hours - by constantly running three ads, split testing the headline and two lines of text - keeping the best one each time and deleting the other two.

Although I improved my CTR sevenfold - my CTR is probably 15 times greater than the average of my PPC competitors.

That means I am getting 15 times more traffic to the site - just as I probably got 15 times more visits to my thread today than any other poster did to theirs.

And split testing the headline is just part of the performance result - there are an infinite number of things you can test on a website to improve the conversion of the traffic you drive there.

So ... in summary - I have proved I get high conversion rates just by posting this thread today.

The first thing with any advertising message is to get ATTENTION ...

And boy did I get YOURS today!

Colin Parker

neild
23rd January 2010, 19:33
I must say Colin, congrats, now your a wizard can you do mine :D

saxondale
23rd January 2010, 20:51
Lets get this into context ...

I did not literally stop my client going broke in 72 hours.

Had he continued with the same lead generation model his business would have continued to decline and HE was worried that could have lead to him going out of business.

In 72 hours I gave him the leads that now make going out of business far less likely - in that sense I have stopped him going broke.

However - the HEADLINE written for this thread 'How - In Just 72 Hours - I Used PPC To Stop My Client From Going Broke' is actually proof positive of the power of split testing.

This thread had nearly 700 views by 6pm today (Saturday).

It is the most viewed thread on the PPC/SEO page today with the exception of Steve Gibson's 'Five Deadly Sins' thread which has been up for four days.

It has therefore had more views than any other thread in this section today - BY FAR.

Had I put up a thread headline like 'A Split Testing Example' I bet the views would not have been more than 50-100.

That is 7-14 times less views than this thread has had.

I put the thread headline up because I know - from testing 1.000's of headlines over the years - which headlines get attention and which ones don't.

And EXACTLY the same applies when you write a PPC ad headline.

That is how I increased my CTR seven fold in just 72 hours - by constantly running three ads, split testing the headline and two lines of text - keeping the best one each time and deleting the other two.

Although I improved my CTR sevenfold - my CTR is probably 15 times greater than the average of my PPC competitors.

That means I am getting 15 times more traffic to the site - just as I probably got 15 times more visits to my thread today than any other poster did to theirs.

And split testing the headline is just part of the performance result - there are an infinite number of things you can test on a website to improve the conversion of the traffic you drive there.

So ... in summary - I have proved I get high conversion rates just by posting this thread today.

The first thing with any advertising message is to get ATTENTION ...

And boy did I get YOURS today!

Colin Parker




in the context of a single interest forum ........... its hardly hard is it.

Tej
23rd January 2010, 20:59
Well it got my attention....

But.. you still have not answered a simple question..post#2 in reply to your opening post. would be nice to reply to that post.. without the gobbledygook.


and now you admit that the 72 hour stuff was BS

:)

I am just a simpleton.. sorry if I am being thick and don't understand the technical stuff...I took your saying as literal... maybe I should have played guessing games..did he, didn't he... maybe I got diverted to the wrong forum and got involved in something way above my head... will know for future reference.

:D

dots and spots Jeff
23rd January 2010, 21:58
Well, inspired by this thread - and with thanks to the OP for the motivation and to Steve Gibson for pointing me in the right direction - I've avoided the temptations of watching Celebrity Britain's Got X Factor Dancing Big Brother on TV this evening and instead set up my first split test - you can only really learn by doing.

I've uncluttered our main landing page* to see if this makes any difference.

Google does make it easy, doesn't it - most of the time spent was deciding on how my alternative page should look, setting up the test (albeit a simple A/B test) on Google only took minutes.

Quite excited by the whole process - I shall report back the results at some point in the future.

Thanks to all for your advice and encouragement.

Jeff

*Our web presence is not huge, and much of that is either direct traffic/referrals so I do worry a little that it might take some time to get a statistically relevant result.

cmcp
23rd January 2010, 22:00
Nice one! :)

SteveGibson
23rd January 2010, 22:31
Quite excited by the whole process - I shall report back the results at some point in the future.

Thanks to all for your advice and encouragement.

You're welcome.

I hope you get a winner. If not, at least, I hope you enjoy the process enough to keep testing.

(you'll get winners soon enough)

Steve

PS Setting up a test is easy.

SteveGibson
23rd January 2010, 22:42
I've just checked a split-test I'm doing for a client.

We started it on the 19th and it's finished already.

A 15.5% increase in conversions with a statistical confidence of 92.8%.

Not quite 500% in 3 days (or whatever Colin's thing worked out as), but 15.5% in 4 days isn't to be sneezed at.

(especially as it was just a small change to the homepage)

Steve

Matt1959
23rd January 2010, 22:50
what is the minimum number of unique? visitors you need in a specific period for split testing to be effective?

traxor
23rd January 2010, 22:51
That is an incredible success story! Really happy for you and your client.

Colin Parker
23rd January 2010, 22:58
Wow.. fantastic.. Zero to Hero in 72 hours.

I am totally naive about websites, SEO, adwords, etc....

But am interested to know ... does leads =actual business? ( income)

The spend might have been reduced to £1000... but its still spend.

Has the income increased dramatically for your client and yourself ( expected income 6 figures etc) to sleep easily... in 72 hours!! ( you get a management fee anyway.. so guess you sleep slightly better)

Really powerful stuff... seems adwords could be the saving of all businesses that are in sh1t creek... and those that are doing well, to be doing miles better.

Why don't all business owners follow this simple solution?

just a thought..

What is your question?

But am interested to know ... does leads =actual business? ( income)
The spend might have been reduced to £1000... but its still spend.

Is that it? How ridiculous a question is that?

If a lead seller (which I am not) sells leads to a buyer it has nothing to do with the lead seller if the buyer doesn't close one lead.

If I work with a client/partner and generate leads with a commission payment on their sales you can bet - after nearly 30 years experience in direct marketing and sales - that I won't get into bed with a sales donkey.

And you can also bet that I will make sure I deliver the highest quality of targeted leads to maximise my commission.

Or is this your question?

Really powerful stuff... seems adwords could be the saving of all businesses that are in sh1t creek... and those that are doing well, to be doing miles better.

Why don't all business owners follow this simple solution?

Yes Adwords could be the saving of businesses that are in 'sh1t creek' - but not yours because you know rock all about it obviously.

I mean - what a daft question. You imply that everyone is a latent Adwords genius who could just turn on the tap and let the gravy train flow. Or that everyone could afford to buy in real expertise and had the cash to do so + the cash to invest in a serious Adwords campaign.

And you try to say I have been avoiding your 'question'.

It's laughable! They are not serious questions that a business person would ask. But there you go - I've answered them.

What next I wonder from you?

And as for you saying I am admitting my reference to 72 hours was BS - go and throw another rattle out of the pram boy.

Colin Parker

SteveGibson
23rd January 2010, 23:01
That is an incredible success story! Really happy for you and your client.

Thanks.

I'm always happy with anything that's 10% or more.

Steve

Colin Parker
23rd January 2010, 23:15
I've just checked a split-test I'm doing for a client.

We started it on the 19th and it's finished already.

A 15.5% increase in conversions with a statistical confidence of 92.8%.

Not quite 500% in 3 days (or whatever Colin's thing worked out as), but 15.5% in 4 days isn't to be sneezed at.

(especially as it was just a small change to the homepage)

Steve

Steve - remember I started from scratch for a client who effectively had no current online campaign.

As you know taking a new campaign from 'zero to hero' is different from taking an established campaign and squeezing another 5% - 20%+ conversion out of it.

However if an 'established' campaign and website is very badly organised and designed getting a five fold increase in leads could well be a doddle - as I'm sure you know.

I've recommended on two members to you who have contacted me via this thread - that's two wee nips you owe me.

Colin

SteveGibson
23rd January 2010, 23:23
Steve - remember I started from scratch for a client who effectively had no current online campaign.

As you know taking a new campaign from 'zero to hero' is different from taking an established campaign and squeezing another 5% - 20%+ conversion out of it.

Yes. My client is successful and had a very high conversion rate already. So there was only so much room for improvement.

(the "cherry on the sundae")

However, I still feel there's a bit more gold in them there hills, so I might suggest another test to him. I have an idea.

I've recommended on two members to you who have contacted me via this thread - that's two wee nips you owe me.

Cheers,

Steve

Scott-CopyandDesign
24th January 2010, 01:12
However if an 'established' campaign and website is very badly organised and designed getting a five fold increase in leads could well be a doddle - as I'm sure you know.

That's why I usually avoid using percentages, as I find they can be misleading (despite being completely correct and useful for examples such as these). If I were to use percentages, I could rightfully claim that I've increased business sales by 2000% and more for some of my clients. However, they only previously had a few sales. It's still an impressive increase, but not as impressive as it may appear to the average person (who may also deem it too good to be true).

Mind you, percentages are fine if you use them with solid sale and profit figures.

Subbynet
24th January 2010, 02:54
A few critical posts questioning why you made such a post, and I don't think its warranted really, he is allowed to keep the customer anonymous.

If I read anything into it, I think Colin has been at this for a while, working hard, and now has landed a brilliant contract, and in the process saved a customer a great deal of hassle so is no doubt very appreciative, and because Colin is happy as punch (and who wouldn't be?) he has decided to share this with us, thank you colin, because many wouldn't. I mean the guy just landed a plum deal!!! Good day's work done (or 72 hrs lol) I think! :D

I'm no expert in SEO, so its good to see the effect a proper understanding of it can have... .

dots and spots Jeff
24th January 2010, 09:19
what is the minimum number of unique? visitors you need in a specific period for split testing to be effective?

Hi Matt - I don't know and is a query I myself have.

However, having set up my own (first!) split test last night as a result of the thread, I eagerly checked my report when I got up this morning. As I suspected I haven't yet had too many visitors overnight, but Google is telling me that in the report : "Collecting data. No high-confidence winner found yet" I think Google works out the confidence levels for you.

I've had great advice from this thread (and forum in general) but I do believe to really find out and understand something you have to have a go yourself - it really was a cinch to set up the test and setting up my own test and reading my own report is helping to crystallize what the theory that I have learnt.

Jeff

Colin Parker
24th January 2010, 09:59
It seems on this thread and others that some people have difficulty understanding the concept on 'split testing'.

Split testing is simply testing different words, text size, fonts, offers, prices, pictures etc., etc., to see what your target customer prefers and wants.

Let's try an example that might help you get the hang of it ...

If you are a guy and I want to know what type of woman YOU find most attractive - I could start by showing you a blonde, a brunette and a redhead.

If you pick the blonde - I could then show you pictures of blondes who are fat, thin and average weight size.

If you pick the average weight size - I could then show you pictures of blondes who are short, tall and average height.

And so it could go on ... and on ... and on.

There are multiple variations of these tests - and I couldn't make assumptions about your dream woman without testing first.

For example - you chose a blonde - but that does not say your ideal woman is a blonde. You might have just preferred the blonde over the redhead and brunette I showed you.

If I showed you a big boobed black haired lass and a small boobed blonde - you might pick the lass with black hair. I would!

So the principle of split testing is actually really simple ...

You are simply trying to find out - by constantly offering combinations of choices - what your target customer prefers.

And there are as many combinations of tests to find the perfect website for example as there tests you could do to find your perfect woman.

So a word of warning - NEVER assume that what YOU like - is what your customer will also like.

Because I guarantee - more often than not - you will be dead wrong.

If everybody liked what you like - they would be wearing exactly the same clothes as you, eating exactly the same food, going to exactly the same places on holiday etc., and ... more worringly (or not as the case may be) trying to steal your lass from you.

Because of course ... every bloke in the world would fancy her.

But don't worry - they don't.

Colin Parker

dots and spots Jeff
24th January 2010, 10:29
Colin is quite clearly an expert in this field and writes with the authority that a great deal of experience brings.

I, on the other hand, am at the polar opposite - whilst vaguely aware of what split testing was, as of 6pm yesterday I had never even looked at website optimizer. Colin wirtes:



So the principle of split testing is actually really simple ...


As a complete newb to split testing I can add that the practice of split testing is also actually really simple (at least at its most basic level).

I watched a 5 minute Google video on Youtube, spent 30 minutes making a slightly different version of one of my pages and another 5 minutes setting up the test on website optimizer - Google really does make it easy!

Its cost me nothing and may gain me something and at least I now understand the whole process a lot more, so if I were to contract someone to do it for me it'd be easier to sort the professionals from the charlatans. Why not give it a go?

SteveGibson
24th January 2010, 10:41
I've had great advice from this thread (and forum in general) but I do believe to really find out and understand something you have to have a go yourself - it really was a cinch to set up the test and setting up my own test and reading my own report is helping to crystallize what the theory that I have learnt.

As a complete newb to split testing I can add that the practice of split testing is also actually really simple (at least at its most basic level).

I watched a 5 minute Google video on Youtube, spent 30 minutes making a slightly different version of one of my pages and another 5 minutes setting up the test on website optimizer - Google really does make it easy!

Its cost me nothing and may gain me something and at least I now understand the whole process a lot more, so if I were to contract someone to do it for me it'd be easier to sort the professionals from the charlatans. Why not give it a go?

I'm glad to see someone getting inspired and setting up a test.

The more people we have testing on this forum, the more others will be encouraged to dip their toe in the water. And, also, the more we can share what we learn from testing.

So, if you update your own site, why not set up a test and see what happens?

As Jeff said, it's easy enough to get a handle on the software. And, once your test is running, you'll see the difference small changes make.

(whether your first test is a winner or a loser, you'll see a difference)

Steve

PS For those with wordpress sites, there's a plugin that allows you to use website optimizer easily with a wordpress site. (haven't used it myself - I'll set up a test with it on one of my sites today - but I've heard it's good)

Scott-CopyandDesign
24th January 2010, 12:10
You might be right Colin. 2010 seems more and more like the year of conversion optimisation by what I've seen so far.

AaronB
24th January 2010, 12:55
Aaron how many of my other 400 posts have you read?

The ones that have got me nearly 200 thanks?

And as for Chris Cardell and myself showing 'how good our marketing is' - what do you us to say - how bad it is?

If I'm about to have an operation I don't want the surgeon to tell me he is an inexperienced amateur who will 'give it a go' ...

The original post was an example of how split testing on Google Adwords can have a massive and virtually immediate impact on a business.

It was put up to motivate people to educate themselves and learn about PPC to help them achieve what I imagine most want - more sales and profits and more money in the bank.

However - as always - the doubters, sceptics and losers can't wait to weigh in with abject negativity. And in so doing they demotivate those who - perhaps like you - are open minded to change and learning.

Colin Parker




Fair do, but it just sounds very American.

Arron.

RBS
24th January 2010, 15:57
Guys,

IMO Colin Parker is pro and we shouldnt doubt it, wasting his time replying to questions from people who feel he is not. His posts have been brainy, he has been thanked 194 times in 97 posts. He is just direct type of talking person, not softy and romantic poster I would say. Sounds a bit army guy, but hey, he has right to do that, because he can make your life a bit better.

Lets just accept he knows what he is talking about and stop writing crap here :)

sm1
24th January 2010, 20:17
Great post Colin - inspired me to split test when our site is up. Not quite sure if people are just jealous or arrogant to believe those results.

As someone said earlier, what sort of unique visitors do you need to get an accurate result? I know the market that I'm planning on entering will have relatively low traffic but I will be split testing direct mail.

Cheers :)

Colin Parker
24th January 2010, 21:37
Great post Colin - inspired me to split test when our site is up. Not quite sure if people are just jealous or arrogant to believe those results.

As someone said earlier, what sort of unique visitors do you need to get an accurate result? I know the market that I'm planning on entering will have relatively low traffic but I will be split testing direct mail.

Cheers :)

Hi Stefan - re visits to get an accurate result - as Jeff has pointed out earlier Google's Website Optimiser will give you a % probability of the 'winning' test.

With PPC ads I normally strike out an ad after 20 clicks each if one ad is outperforming another by a factor of 2. Less than that I might run the ads for 50 or more clicks each.

Colin Parker

sm1
24th January 2010, 22:24
Thanks for clarifying that :)

freshpurple
25th January 2010, 11:01
With regards to Steve's previous ponderings over why more people don't use split testing, I suspect that technical ability probably has a large amount to do with it - even after you have grasped the idea (and, as many people have pointed out, it's a relatively simple one) you have to be able to produce two different versions of a web page.

If you're someone who doesn't know how to code html this is quite intimidating (and it could be worse if you are using an e-commerce software package). In which case you have to go back to your web designer every time you want to organise a split-test, which costs money, takes significantly more time and can be quite a hassle as well depending on your designer. I also imagine managing this kind of campaign by proxy is not ideal.

(Of course, perhaps this should be the designer's job in the first place, but that probably requires a deeper professional relationship than the average pay £1XXX and I'll give you an e-commerce website)

SexyThing
25th January 2010, 13:07
PPC is for amateurs that don't understand SEO:rolleyes: that's why I am learning SEO

and this crap about changing 2 words within the whole website and incerasing the orders by 30-40% pleasee:rolleyes:

the only thing that is going to increase the orders that much is putting the price down. I do not care how the product is explained if it is £400 on your website and £350 on another website guess where I am going to buy from;)

RBS
25th January 2010, 13:13
PPC is for amateurs that don't understand SEO:rolleyes: that's why I am learning SEO

and this crap about changing 2 words within the whole website and incerasing the orders by 30-40% pleasee:rolleyes:

the only thing that is going to increase the orders that much is putting the price down. I do not care how the product is explained if it is £400 on your website and £350 on another website guess where I am going to buy from;)

I am very sorry, but are you so stupid or you just pretend to be? You better keep trying to sell you ass for 500£ a day to some big blokes and stop talking crap. You know nothing about this stuff :rolleyes:

Scott-CopyandDesign
25th January 2010, 13:28
I am very sorry, but are you so stupid or you just pretend to be? You better keep trying to sell you ass for 500£ a day to some big blokes and stop talking crap. You know nothing about this stuff :rolleyes:

I wouldn't waste your time on this guy. Just completely ignore.

Pete Crane
25th January 2010, 17:02
PPC is for amateurs that don't understand SEO:rolleyes: that's why I am learning SEO

and this crap about changing 2 words within the whole website and incerasing the orders by 30-40% pleasee:rolleyes:

If you really are claiming to be learning SEO, you will eventually realise that at some point making such a change will have a similar effect. Or more.

the only thing that is going to increase the orders that much is putting the price down. I do not care how the product is explained if it is £400 on your website and £350 on another website guess where I am going to buy from;)

So, your sole purchasing motive when buying online is price and nothing else? If that is the case, why are you claiming to be learning about SEO? Surely this is irelevant to you as you won't care about rankings, you're only bothered about price.

Colin Parker
25th January 2010, 17:16
PPC is for amateurs that don't understand SEO:rolleyes: that's why I am learning SEO

and this crap about changing 2 words within the whole website and incerasing the orders by 30-40% pleasee:rolleyes:

the only thing that is going to increase the orders that much is putting the price down. I do not care how the product is explained if it is £400 on your website and £350 on another website guess where I am going to buy from;)

Amatuers wait until they get on Page 1 of Google to find out they are targeting the wrong keywords ... and their site doesn't convert.

Professionals test what converts first on PPC ... and then start optimising for SEO.

So where does that put you Mr Sexy Thing?

In the cheap minded, clueless amateur bracket.

Colin Parker

SexyThing
25th January 2010, 22:15
you know nothing about me, only thing you know is I am sexy;)

I am not clueless, lets see who is going to rank in top 3 in 3 months for the term model agency;)

Amatuers wait until they get on Page 1 of Google to find out they are targeting the wrong keywords ... and their site doesn't convert.

Professionals test what converts first on PPC ... and then start optimising for SEO.

So where does that put you Mr Sexy Thing?

In the cheap minded, clueless amateur bracket.

Colin Parker

sm1
25th January 2010, 22:22
Not you? :|

SexyThing
25th January 2010, 22:31
me me;) and also I am testing the SEO on another website by simply outsourcing the work to India and see if I can rank for a term for cheap as chips price;) and before you say anything the company I am employing got 562 positive reviews on the freelancer website;)

GreatSEO
25th January 2010, 22:38
Unfortunately reviews do not mean a thing. SEO in its basic form is a very simple process and if you use Google you will find all you need.

Using a company in Indiawill lead to all sorts of problems one of the biggest being the communication in English breakdown if you are targetting UK individuals.

The more complex stuff is what will take you years to test and measure and will without a doubt be something that most people give up on in the long run.

Try some of the stuff over at seomoz I am sure it will give you a good start

Hope that helps

Dave

An Oasis
25th January 2010, 23:16
Amatuers wait until they get on Page 1 of Google to find out they are targeting the wrong keywords ... and their site doesn't convert.

Professionals test what converts first on PPC ... and then start optimising for SEO.

So where does that put you Mr Sexy Thing?

In the cheap minded, clueless amateur bracket.

Colin Parker

Colin don't engage, it's a nutter escaped with no meds...don't waste your time.

papverpoppies
25th January 2010, 23:37
Its DWP.

Another name, but the same format.

Girls chasing him, top ranked site/business in x amount of weeks, spelling errors galore.

Need I go on.;)

Poppy

SexyThing
26th January 2010, 08:41
yes I am the Department for Work and Pensions!:rolleyes:

(someone have not taken their medicine today)
Its DWP.

Another name, but the same format.

Girls chasing him, top ranked site/business in x amount of weeks, spelling errors galore.

Need I go on.;)

Poppy

Scott-CopyandDesign
26th January 2010, 11:13
SexyThing (http://www.ukbusinessforums.co.uk/forums/member.php?u=60646)
Banned


Best thing I've read all day.

*Lexxy*
26th January 2010, 12:10
you beat me to it Scott! :D

1weekSEO
26th January 2010, 14:29
Where have I been?

This has got to be one of the best threads I have read in years - not just in business forums but for a while I had no life out of a DJ forum in my earlier years - this was quality!

I especially liked the 'banned' posting ..

Going back to Adwords (if I may) I test ad campaigns and have a couple of adverts with different text and delete the ones that don't perform as well. When I joined this company last year, they only had one ad campaign he had something like 220 keywords / phrases and one advert. If there's anything else for me to learn from Adwords I don't know what it is, but then again, "I don't know what I don't know"

But what I do know is that with a £4,000+ monthly Adwords campaign this gives me room to test on a weekly basis, analyse and update. Thank you to all contributers, I wish I could split test but as someone else pointed out (at present) I can't get any extra webmaster time.