Enquiries written by AI

[tony84]

This week I have had 2 enquiries from real people come in but they are written by AI.

I had never considered this before but as a small business I like people writing their own messages. I always remember some training years ago saying you should mirror and match the person you are speaking to. I thought it was BS to be honest, but I have realised I do actually do it subconsciously. I cant do that when replying to a robots message.

The message is asking questions which the person has been told are important by AI but after 15 years of doing the job, those questions are rarely questions that crop up. It means people are asking questions they think they should be asking rather than questions they probably want to ask.

I dont get it. I dont want to say it is lazy because I imagine it involves more work than just writing a message.

 

[Mark T Jones]

We are in slightly different environments, but my reaction to any enquiry that appears to be automated or non-specific is a polite version of.
We are very much a people business, not a machine. If a decosion-maker would like to discuss your requirements, please can they call me on ....

Sometimes they actually do!

 

[tony84]

Fair.
I have people who work for me now and I am not sure they would like me potentially losing decent enquiries.
But also, I think it is just the way the world is going, I just dont understand it.

 

[BusterBloodvessel]

Yes, we get it regularly now and it irritates the hell me. On the one hand, I get that it may be down to language issues and people making sure things translate well.

But often it's because people are completely inexperienced so they're asking ChatGPT to tell them - exactly as you said above - what it thinks is important. Which then often means all the stuff that's irrelevant at this stage, and/or makes it sound like they want me to do their job for them.

This is one I had last couple of days;

Hello,

I am currently setting up an e-commerce business in **, focused on practical products for car owners and pet owners .I am interested in your pet hair removal products, particularly those designed for use on car interiors, and I would like to learn more about your wholesale and private-label options.

Could you please provide me with:

- Wholesale/B2B price list
- Minimum order quantities (MOQ)
- Available products suitable for the ** market
- Shipping options and costs to **
- Lead times
- Private-label possibilities and requirements
- Product samples and sample pricing
- Product specifications and available certifications/documentation
- Your best-selling products in this category

At this stage, I am particularly interested in starting witha small test order before committing to larger quantities.

It's like wooooahhh... slow down there! He/she has barely even narrowed down the products they want to sell yet they want to know about certifications/documentation! Depends entirely on the product(s). Asking for our best selling products is basically "tell me what I should sell". Lead times, MOQ's etc again all depend on product.

It's just 'head first' in rather than an exploratory conversation to try and understand what they're looking for, where we can then help and then take it from there. Now of course all of the above we can supply and help with - but not from a random outreach enquiry through our website.

And don't get me wrong - I have absolutely nothing against people starting out in business or if people are inexperienced. I have often had lengthy chats with people who are starting out or looking at possibilities, knowing that there is a 99.9% chance it will never go anywhere, or at least nowhere that would get us sales! But if people come on and say "I'm looking into selling on xyz, just wondering if I could chat about products abc that we're interested in for our own brand" - well that's more likely to start a decent conversation with me than demanding all of the above.

The type of email above quite frankly receives a polite, short reply along the lines of the "yes we can do all this but we need to understand your specific products/ranges of interest, and your business model and expectations in terms of delivery times & quantities". I politely point out that MOQ's typically start from 500 pcs and everything is manufactured to order but we would be delighted to discuss further once we know more details from them.

Unsurprisingly, it's very rare that we ever hear back.

 

[Frank the Insurance guy]

But often it's because people are completely inexperienced so they're asking ChatGPT to tell them - exactly as you said above - what it thinks is important.

^^THIS^^

We are getting more and more enquiries that are AI suggestions - I get that Business Insurance can be complicated for a start-up that is not familiar with it, what to ask for and what it does!

My response is tailored to suit the enquiry as I would normally- if a start up, likely they need hand-holding through the insurance process, or if established then likely to be more bespoke. In both cases, my response would usually respond to the AI's specific query/suggestions, including sometimes challenging them! - As the enquiry is AI driven, I take the assumption that my response will just be thrown back into AI to decipher for them - AI is quite often wrong when it comes to some of the technical aspects of insurance and therefore challenging it shows the recepient that we know what we are talking about (Their AI should confirm we are right!!).

I think its just something we all need to get used to. Client's are going to look for AI to support them more and more so these types of enquiries are only going to increase.

 

[Newchodge]

Try advising an employer how to deal with an AI generated employee grievance. What used to be a few sentences along the lines of "My boss, Fred, told me off because I was 10 minutes late, when he knows I have to take the kids to school first" is now 12 pages of waffle, spurious caselaw, irrelevant legal argument and threats if acceptable action is not taken yesterday. An employer does not have the luxury of rejecting the drivel, but has to deal with it.

 

[BusterBloodvessel]

Try advising an employer how to deal with an AI generated employee grievance. What used to be a few sentences along the lines of "My boss, Fred, told me off because I was 10 minutes late, when he knows I have to take the kids to school first" is now 12 pages of waffle, spurious caselaw, irrelevant legal argument and threats if acceptable action is not taken yesterday. An employer does not have the luxury of rejecting the drivel, but has to deal with it.

Good god yes, I can imagine that's an absolute nightmare! As we get it sometimes with product complaints. What used to be "this is faulty, the light won't turn on" now turns into them telling us our legal obligations for replacement, expected product lifespan, suspected failure reason and then demanding our response within 48 hours or they will escalate to legal action!

 

[DontAsk]

I've been getting those sorts of enquiries for years from people who clearly have not looked at my companies website. AI makes it easier, I guess.

 

[Frank the Insurance guy]

I think the biggest problem is that some people are starting to believe the AI Outcome and advice over the professional who has decades of real world experience!

 

[tony84]

I think the biggest problem is that some people are starting to believe the AI Outcome and advice over the professional who has decades of real world experience!

If you say dont trust AI which is the thing they have been trusting you sound like an over the hill bitter old man.
You have to be diplomatic about it and it is becoming a job in itself.

I use AI for some bits but I also go off and check what it is telling me as I know from experience its not necessarily wrong but it is like a politician or conspiracy theorist - only half the truth.

It like that people are trying to educate themselves beforehand. I have always been in support of that and there are times where customers teach me something new, I love that!

But I dont like people not asking questions they think are important, I dont like people being given incorrect information/half truths from something they trust as it is awkward when trying to correct it and I just dont like AI writing emails to me as you cant get a sense of who you are working with. I like to know my customers.

 

[antropy]

We are seeing it crop up as well, not just in enquiries but also sales outreach that enters our inbox. Sometimes it takes a minute to skim the email to parse whether they want our web dev services or are trying to sell web dev services to us!

It's also quite likely that some of these people are asking their AI agents to search for services, then in the same chat asking it to generate an enquiry email for them to send to the companies that pop up.

- Kelly at Antropy