My AskAI · Mike Heap & Alex Rainey

From Corporate Ladders to AI Pioneers: How Two Entrepreneurs Turned a £1.2M Setback into a Revolutionary Customer Support Tool.

February 11, 2024
Share this story

Table of contents

  • Mike Heap and Alex Rainey
  • London, UK
  • Started in 2023
  • $325k ARR
  • Bootstrapped
  • 18,000 monthly visitors
  • myaskai.com

What's your backstory?

Both Alex (my co-founder) and I started our careers in the big corporate graduate schemes of Accenture and EY in London, respectively.

He was in tech consulting and I was in the enthralling world of Banking and Capital Market audit (accounting, basically), the obvious next step for someone with a Maths and Physics degree.

Alex spent 6 years at Accenture before he left to start his own travel InsurTech startup called Pluto which managed to raise £1.2m and get a load of users before the COVID-19 pandemic unfortunately wiped it out.

I spent 9 years at EY, moving from audit to an "innovation" role (as fluffy as it sounded) building a team of data scientists, robotics, and blockchain engineers to try and bring audit to the 21st century.

But it quickly got to the point where I couldn't take the bureaucracy and slow pace anymore and needed to get out.

In my last few years at EY I had been working part-time on an app to find people to play sports with alongside a mental health app so thought I'd take the plunge (stupidly in hindsight) to quit with just 3 months' savings to pursue my entrepreneurial ambitions.

Very quickly I realised that was a bad idea and needed to find work, but given everyone just saw "accountant" on my CV it was very hard to land a product role.

Luckily, from a placement I had had at a FinTech while at EY I was able to tap up an old friend who had moved to be CTO at a renewable energy startup and was looking for a Project Manager. So I took the gig and contracted there for a while, before moving to an InsurTech scaleup, all the while trying out different ideas to see what might stick between each contract.

In late 2022 a mutual friend introduced Alex and I to one another, namely to share experience on how to find contract Product Management roles. We met up and ended up just talking about AI, indie-hacking, a side project Alex had just built and sold in 2 months and realised that building something new together with AI would be way more fun than contracting.

What does your company do and how did you come up with the idea?

One of the AI ideas we ended up launching (as part of one of Ben's Bites hackathons in Dec 2022) was the "No Code AI Model Builder", a super-catchy name for something that allowed non-techy people to make fine-tuned versions of OpenAI's models.

Built and launched in 3 days due to us both being ill the week of the hackathon, we put it out on Product Hunt expecting crickets.

But it got noticed, and ended up being like #10 product of the day I think, getting 2k visitors on our basic site and even our 1st paying users!

It remained busy over the holiday period and early Jan and we started planning new features but found that while we were still making sales, whenever we spoke to users they rarely could be bothered to do the upfront work required to create a fine-tuned data set, they wanted to dump in data and get something out the other end they could ask questions to.

So as we saw more people building "chat with my book" type things we realised we needed to make the platform to allow anyone to do it. We made a simple screen in our current product asking people if they wanted to pay $99 for a pre-sale to use this product that didn't even exist.

$5,000 later, realising we had proven demand we knew it had to be built.

3 weeks later (and a slightly ruined holiday for Alex in Thailand), we launched My AskAI to the world.

My AskAI started off as a tool where you could upload a variety of content - from PDFs, Docs, websites, Google Drive, Notion etc into the AI and then "chat"/interact with it as if you were talking to someone.

People used it for a whole bunch of use cases - research, knowledge bases, characters, but the primary use case we saw was Customer Support.

So we doubled down on this use case and now the product is built around that:

You add your domain and we build an AI Customer Support Assistant around your content that you can add to your site, if it can't answer it passes you over to a human. If you want to know what people are asking, we use AI to generate insights and aggregate conversations.

The main purpose of the product is to enable SaaS businesses to move from answering the huge volume of generic support requests they get to focus on customer success and the tougher-to-answer queries, getting actionable feedback along the way that allows them to improve their product.

How did you get your first 10 customers?

Our first 50 users all came from pre-sales from our previous product (the users for which came mostly from our Product Hunt launch and a little from building in public on Twitter).

What steps did you take to understand that your product was a good match for the needs of your target customers?

Post-launch, we would offer every paying user a set-up call, that a decent number took us up on, we would treat these as user interviews to understand what they wanted.

Part of the issue was, that we did things a little backwards in that we were selling the features and letting people figure out what to do with it, this is generally a bad idea, but initially the technology was so new and there was so much excitement that people were very happy to just tinker and see what it could do.

As the months went on we started to see patterns of usage and how they wanted to use it, so we tried to focus on building more of those features, around customer support and internal knowledge systems for businesses.  

But we still had a long tail of users wanting to use it for other things and so this made a lot of things more difficult - prioritisation, marketing, support.

We were super reactive to customer requests and bugs, manned our own support and would generally fix things within minutes or hours of a user raising them.

In the summer, the momentum started to slow as a proliferation of similar tools made it much more competitive and drowned us out with noise, also because of the broad applicability this made it really hard to find repeatable and economic distribution channels.

In October we decided something needed to change and we dug heavily into all the data we could get our hands on from our business and asked "how are most people using this", the answer was pretty clear - customer support.

This then led us to re-think the product - if we were to rebuild it from the ground up for this use case, what would we do, it was quite freeing to have that level of focus.

We then spent 7 weeks re-building the platform from scratch to what you see today and are very happy with the result!

How did you reach and acquire your first 100 customers?

Our first post-launch (i.e. not pre-sale) users came from a new Product Hunt launch, plus things like:

- Replying to big AI influencer accounts on Twitter with our product
- Getting featured in AI newsletters when the hype was at its hottest
- AI directories featuring us
- Continued building in public on Twitter

We were effectively just targeting the early adopters, pretty much agnostic of use, it's not a strategy I'd recommend in pretty much any other situation, but I think it was the right one for that particular moment.

What distribution channels did you try that didn’t work?

Oh man, we have tried it all, and I would caveat with a couple of things:

1. We probably didn't try a lot of these for long enough that we could definitively prove success or not
2. AI has been weird because so many of the "marketing" methods early on were just growth hacks that worked for a period before becoming unviable very quickly
3. We didn't have a clear positioning when trying any of these, which would have made conversion a lot less likely anyway.

That said, here is everything we have tried that we are no longer doing:

- Google Ads - we hired an agency, but we couldn't get a low enough CPA to justify.
- Replying to AI influencers on Twitter/LI - worked for a while then got noisy and influencers got less engagement.
- Sponsored Reddit post - worked in the hype phase, less effective now.
- Paid newsletters - again, worked in the hype phase and sometimes for a 1st placement, but not so well as an ongoing channel.
- Paid TikTok influencer - Only tried once so not a good test but didn't bring anything and just didn't feel it was a good fit for our product (B2B mainly).
- Paid Twitter ads - Not sure we got a single paying user.

What specific tools, software, or resources have been most helpful in growing your business?

  • ChatGPT - without doubt, for coding help, brainstorming, analysis and research.
  • Twitter - we have learned so much from the Indie Hacker community, wouldn't have been building what we have without them.
  • CarbonAI - they have been phenomenally helpful for us, allowing us to focus on the value bits of our business instead of the data side.
  • Browserless - similar to Carbon they have been a key part of infrastructure since early on.
  • Bubble - we wouldn't have done any of this without it, it allowed us to move super quickly and there isn't anything we have found so far we haven't been able to build on it.
  • CustomerIO - great email platform with a good deal for bootstrappers, way better than SendGrid.
  • Senja - for testimonials, hard to prove as we had from day 1 but think it definitely helped us standout from a lot of tools out there as more legitimate.

Who are some recommended experts or entrepreneurs to follow for learning how to grow a business?

(All on Twitter)

Andrea Bosoni
Marc Louvion
Sveta Bay/Dan Kulkov
Damon Chen
Andrew Gadzecki
Pieter Levels
Danny Postmaa
Olly Meakings/Wilson Wilson
John Rush
Arvid Kahl
Tim Bennetto
Tibo

How did you make the transition from a side hustle to full-time entrepreneurship?

I tried a few different ways:

1. I took a sabbatical from work for 3 months years ago, it wasn't enough time and I didn't know enough to make it work

2. I went part-time 2/3 days per week and worked on my ideas on the other days, didn't really work as the day job just consumed everything

3. I started contracting, working for 6 months, then having 6 months on projects and ideas, this worked well for me, it gave me a deadline, but was hard finding contracting work

What drives you to do what you do?

I realised last year that I just love solving problems and building slick experiences. It means that even though customer support isn't "sexy", when I had the freedom to think about it in a more first principles way, you can come up with some pretty neat ideas.

I was originally very mission-driven and would still love to do something more meaningful one day, but think that you can also find deep meaning in most things, it's just about perspective -

We save people time and remove frustration, which gives them a better quality of life AND allows them to focus on more interesting things, the impact of that is hard to quantify really,

I am also very strongly motivated by freedom, I am not sure I could go back to working for someone now, so kind of necessary that I make this work!

Any quotes you live by?

Ha, bit grand to say I live by them, but I always liked these (and hope these are the right people...):

- Be the change you want to see in the world - Mahatma Gandhi
- Great minds discuss ideas; average minds discuss events; small minds discuss people - Eleanor Roosevelt
- If you can't explain it simply, you don't understand it well enough - Albert Einstein
- Only put off until tomorrow what you are willing to die having left undone - Pablo Picasso
- If you're going through hell, keep going- Winston Churchill
- The biggest risk is not taking any - Mark Zuckerberg

Your links + socials

Linkedin

Mike Heap Twitter

Alex Rainey Twitter

Alex Rainey Linkedin

My AskAI

Share this story
Back to all stories