Ask HN: Have any of you left a startup that was selling snake oil?

58 points by Dr3vil ↗ HN
Hi all,

I am in an interesting situation. I thought I'd ask you all to see how you've handled similar situations.

A few months ago, I joined an 'AI' start-up. Most of my experience lined up well with the company vision, and I thought 'why not?'. They gave me what I was looking for in terms of salary and have funding to run for the next year.

The company is really just a consulting firm that deploys custom solutions for clients: this is totally fine by me, because it's really interesting and fun. But, the CEO is going to investors pitching the fact that we've built/are building some deep learning engine that can be applied to all problems. Investors are starting to do due-diligence on us, and I've repeatedly told him that there is no 'engine'; we're a consulting company. All problems we face are unique, and we provide a solution. There's no magic ML algorithm that works for all problems. This logic falls on deaf ears.

I feel that it's best to move on from here eventually - I may give it another few months to see what happens.

Any of you have any interesting takes/similar stories? I'd be curious to hear how you handled it.

45 comments

[ 73.8 ms ] story [ 441 ms ] thread
What you are describing doesn't sound so far off from say "IBM Watson" ;)

There are so many companies doing real work in AI. NVidia for example. Or Waymo. You could even start a GPU Cloud services business. Or seek an advanced degree in academia. Best of luck!

Worked out OK for the Autonomy founders as well, according to HP.
It doesn't sound like snake oil. It does seem like you are actually building AI solutions, and guess what, you said it yourself:

"this is totally fine by me, because it's really interesting and fun"

The founder needs to do what he has to fund the dream while it is still amorphous. Once you've done enough consulting, hopefully you will be able to re-use some of what you've done over time, and start focusing on a single problem area and building that 'engine' - that is, if you stick around long enough for it to be realized (and dont run out of money first)

> The founder needs to do what he has to fund the dream while it is still amorphous.

Meaning, lie about the product? Maybe this is normal; I'm not part of that world, so I wouldn't know. But hiding it behind fuzzy phrasing seems like doubling down on dishonesty.

Agreed. If a founder is willing to be dishonest to investors, he/she are also likely to be dishonest to customers and employees. There’s a fine line between marketing fluff, which is ugly but acceptable, and outright lies, but if you feel uncomfortable, it’s likely a problem.
(comment deleted)
Isn’t this kind of what a startup is? In the early 2000s I worked for a 5 man band company that rode the pre-smartphone mobile wave to a 500 million pound valuation and 400 employees, sounds cute now I know, and our first days were similarly seat of the pants.

Our CEO could sell anything, and then we built it.

If you’re not defrauding customers but leaving them with a useful solution, where is the snake oil?

Sounds like the engine is you guys :)

> where is the snake oil

Doesn't the original post describe a CEO lying to potential investors about a product that doesn't exist? Maybe that doesn't map perfectly to the 'snake oil' analogy, but it's close enough.

Not everyone is comfortable with dishonesty. If it's a necessary part of doing business in this field, fine, maybe the OP needs to hear that. But lies are lies, even if sometimes they work.

isn't deep learning engine a set of scripts to solve some problems anyway? Are they trying to build PyTorch and TensorFlow? If they are proving solutions for customers, it does exist. Sometimes developer don't see the value of things because they are dealing with it everyday. For us it is scripts, for other people, it could be magic.
No. OP has described a liar. A truthful version would be "we don't have this but my team can build it". This is not what's happening.
I doubt Microsoft has achieved the deal with IBM with the truthful version about "their" OS.
Quoting OP:

> the CEO is going to investors pitching the fact that we've built/are building some deep learning engine that can be applied to all problems.

> If they are proving solutions for customers, it does exist.

From the perspective of a client with a particular project, provided the client's needs are met at a reasonable price, they are happy. It does not matter if their needs are met by application of some state of the art proprietary general purpose AI tech, or by application of standard open source technology and a lot of hours of consulting/professional services work.

From the perspective of an investor, there is a huge difference between a company that has state of the art proprietary AI tech, and a company that merely offers professional services using the same tech than any competitor has access too.

The former company has a much bigger competitive advantage / "moat" and is likely a far better long term investment.

Claiming that you have proprietary tech to potential investors when you don't is fraud.

Claiming that you are planning to build such tech is perhaps not fraud if you are indeed planning to build such tech, it'd be up to the investor to evaluate how likely that endeavour is to succeed.

Actually, I agree.

Pitching this to investors is fraud if you don't have it or are not planning to build it.

I misread, and was more thinking of it from perspective of a customer buying a solution, and getting an AI solution in the end.

I would leave sooner rather than later. Maybe read a few articles about the Theranos debacle if you can't readily grasp why.

Lying to investors goes by an ugly F word: Fraud. It can involve jail time and it can hint at a willingness to do worse things.

/personal opinion, not based on actually having worked for such a company

Right. Lying to users is one thing; growth must be prioritized or the business might fail so lying to customers is excusable. But, lying to honorable(i.e. wealthy and charitable) investors is morally corrupt.
The OP only mentions misleading investors, and my impression from the OP is that the clients themselves are getting the consulting services that they're being sold on.
without judging the original situation, misleading investors is bad. Customers know what they are getting and can accordingly choose to extend or cancel the existing business relationship. That's easy. Investors rely on management's communications to understand how their money is being used - If management is misrepresenting, how are the investors supposed to know if they should invest further in this company? Shit like this is the entire reason the SEC was founded.
It isn't reasonable to infer that my statement that lying to investors is fraud somehow means lying to customers is totes okay.
I think it is CEO's job to keep investors hope high. A engine doesn't work like magic anyway, you are always going to customize it. Thinking TensorFlow as a engine. I don't see anything wrong with that. Anything can be boring, and CEO has to make it exciting. I have this experience with my recent project that I am in charge of, the team facing a lot of doubts and pushes because I didn't draw the picture sexy enough for management. CEO takes on the task of making sure investors are happy to keep the company going and making sure team members are happy as well. What you see is 80% of the value of your company, and the CEO sees 120%, that's the difference and it is the right thing to do.
Lying to investors is not okay, and OP describes a CEO who they believe is flagrantly lying.

This is a kind of "excitement" no company needs.

> This is a kind of "excitement" no company needs.

Citation needed? It would be good to have a (probably impossible to perform) study that compares the performance of companies where CEO is lying vs the ones where they're only telling the truth. From my personal experience, I've seen a lot of success built on lies.

It's questioning the ethics of lying, not the production advantages of being unethical.

The two of you have different priorities. You are both right and both wrong.

It's not really the kind of thing that requires a citation. However it's trivial to Google and see there are a whole range of opinions on the matter.

So, lies and deceit to increase investor confidence?

Wasn't this a big factor in the financial crash?

Doesn't seem like a good plan at all.

You seem to have a capable CEO that is able to bring in the funding. Unless he is blatantly lying I'd say what you are experiencing is pretty standard. Throwing in some buzzwords is not selling snake oil. It used to be "big data", now it's "AI", next it'll be "blockchain".
> the CEO is going to investors pitching the fact that we've built/are building some deep learning engine that can be applied to all problems.

> there is no 'engine'; we're a consulting company.

This is pretty close to fraud. There's a huge difference in return to investors between skilled data scientists doing one-offs, and an engine that does the work on its own. If they're not even trying to build a generic AI engine, it's over the line.

> Unless he is blatantly lying I'd say what you are experiencing is pretty standard.

I would say that he is lying, if the description of the situation is accurate, and that "what you are experiencing is pretty standard" isn't _really_ true. But it's not far wrong either. And that's unfortunate.

> There's a huge difference in return to investors between skilled data scientists doing one-offs, and an engine that does the work on its own

This is kind of the crux, right? If the investors believe that there is a future roadmap to essentially automate what the team is doing ... That's one thing. If they are investing today because the CEO is saying "Look! It's automated!", then that's pretty much on them.

That said, these investors have literally millions of dollars to complete due diligence, and if they miss on something like that, it's on them. Investors are putting so much money into these companies on "demos" and without really looking under the hood. So if they complete their due dilligence, their essentially saying "we are investing in your current product and our current understanding of it". I don't think there's going to be a fraud lawsuit here.

Theranos was changing the output data that the investors saw, not just saying "we do blood tests better!". That was straight up fraud because the data reviewed in due diligence was false.

More importantly, an AI engine is IP that could make the company an attractive acquisition target at a higher price.

Talented consultants not so much, because they may not stick around after the acquisition.

If the CEO doesn't deliver something either (a) very much like an engine that can scale the business without hiring data scientists for every job, or (b) some other very exciting thing, he's going to have a tough time.

Maintaining investor confidence is almost as critical to a CEO as raising money.

I detect that from the beginning now. In your case it can't become big because consulting doesn't scale. As they are into AI, they won't even think at trying to scale consulting.

Next time, negotiate nothing less than cofounder at least to get a control over the purpose of the company.

> consulting doesn't scale

IBM disagrees

>But, the CEO is going to investors pitching the fact that we've built/are building some deep learning engine that can be applied to all problems.

Are you directly involved in the meetings the CEO is having with investors? Do you have first hand knowledge of how the investors are reacting to what the CEO is saying? Is what you are saying based on your experience in those meetings, or from other conversations with your CEO? Remember that your CEO may have conversations with you that don't, in any way, reflect the conversations they are having with investors.

The world of business and money is filled with hype, painting pictures, "vision for the future", etc. Startups, Big and Small Tech Companies, Non-Tech Companies, it's all the same if you're trying to get VC money or budgets. What's comfortable for CEOs,VCs, and marketing may not be comfortable for somebody in a technical role. The people investing the money or approving the budget need to do their due diligence properly.

Rather than asking about your CEO, ask yourself if your own sense of morality is being violated. I've been in a similar situations before, where the company I worked was using consulting hours to sell a terrible (in my view) Data Science product. The "Product" did more to satisfy the person signing the purchase order than anybody who actually had to use it. That being said, all the consultants were sincere and the customers were really happy with the work we did, even if they had to buy that terrible product to get more time from the consultants.

If you are enjoying the work and doing right by your customers, then let yourself enjoy it. If your CEO is saying you are doing cutting edge cancer research with your product and you are running the models for a active clinical trial, find another job!!

If the business seems to be doing well and growing, I’d reconsider if what he’s doing is really that bad. This is the reality of “AI”/ML - there’s no magic bullet, but naive people want to be fleeced into thinking there is one.
Your story sounds too familiar. I worked at a snake oil startup for almost 2 years, essentially a "Theranos" for food (google "vegan mayo scandal" to get the incredible whole story).

The CEO dramatically changed the focus and pitch of the company on a dime multiple times just to impress potential investors. The pitch at many points included "AI" (scientists using R to run regression models). Later it evolved into claiming we'd be in the "cellular agriculture" space before anyone else. At best, he was exaggerating and extrapolating minor discoveries. At worst, it was full blown fabrication, total vaporware. And to make things even more frustrating, he was brazenly uninterested in the science (and the input from the actual experts in the room). Everything was about a compelling sales pitch and continuing to buy the company time and funding.

I watched multiple waves of incredibly talented scientific and technical staff be sold by the vision then run for the hills after a year or two of working there. My advice is - if your ethics are being compromised, get the hell out of there.

I googled "vegan mayo scandal" and here's a quote from the first result - "He notified Hampton Creek’s two outside board members, from Khosla Ventures and Horizons, that he suspected Tetrick was misrepresenting the company’s finances. Both firms stuck by the CEO, and the board launched no formal investigation."

The quote is from a Bloomberg article - https://www.bloomberg.com/features/2016-hampton-creek-just-m...

It's just astonishing to read that an company executive told board members that something seemed wrong with financials and the board members didn't immediately look into it. If the people who are supposed to be doing the due diligence aren't doing it, who do they think is going to do it?

I was at a party a while back. A friend has friends whose families are all pretty wealthy. Friend asks an old buddy what he is doing since college. He responds "I'm a board member now!... Dad says I just have to not ask any questions."

I thought it was a joke, it was not.

If only they made "Blockchain-AI-Vegan-Mayo"...
Get out! If you don't feel comfortable with CEO, then it will be very hard to feel good about your work.

A while ago, I worked for a consulting company that specialized in e-commerce with fulfillment. Soon I realized that their clients are mostly MultiLayer Marketing companies. MLM is legal but very unethical industry. Our clients referred to their customers as idiots or bored housewives. Some of my coworkers tried to quite their concise by convincing themselves and me that these people are lazy who are trying to get rich quick. So they deserve to lose their money. They would say it is a cheaper than business school.

I tried to tell myself that I am just a worker, just following orders. It was a good job with a lot of benefits but it was very soul crushing. Leaving that job was the best ever thing. My mental health improved the minute I handed my two weeks notice.

I think what you'll find as you get further in your career is that sales-orientated individuals (revenue sales or investment-as-sales) tend to over-represent their capabilities to the people they are selling to. If you quit every company where you find that "sales is selling things that we can't build" there will not be many companies left to work for.
Your CEO's job is to build vision and hype. He's supposed to look toward the future - the eventuality of what you guys are working towards.

This is so facepalm/cringe and typical of engineers who think they know better and everyone is an idiot lol.

Theranos agrees!
This is lots of companies I've worked at. Producing vaporware solutions for problems that may or may not be defined via user requirements (if defined at all). I'll just tell you about the worst one I worked at:

Under the hood, the company was down to 4 engineers (we sold the company as a team of 30 engineers). We were basically doing Ponzi-style SOWs promising time to clients over time another had paid for, etc. Lots of our 'solutions' were just white-labeled third-party software bundled up into a shitheap. We didn't have any actual IP because all of the stuff we made was licensed by the hosted providers we used. We actually stopped really acquiring any new tech or updating or methodology maybe a year into me working there. My boss shifted focus onto social media (blog posts about cloud shit, tweeting about cloud shit, facebook about you guessed it--cloud shit.) He even went as far as to spend ~20,000 on high end production and recording/streaming equipement and software to start doing live webinars about more, yes, cloud shit. The content of the webinars was whatever the engineers could manage to scrape together at 2am after finding some time in their massive stack of actual work to do. Our boss never learned how to set up the shit, the webinars barely ever streamed successfully, and the social media best I know never made a mark.

I've been gone a while but last I heard the CEO is still hellbent on being the most widely visible adopter of the cloud in the world.

I was hired on to a company in midtown looking to build out an HPC environment for analytics. I live pretty low in the stack, so my time was spent head down building a system with filesystem and network performance to the lead dev's spec.

The CEO had experienced a major exit in dotcom v1, and was funding the new venture himself. Lead dev was one of CEO's generals from the bubble win. Money was basically unlimited.

This meant we were building a hoss of a system with a latest-gen infiniband fabric, the most ram per-core Intel could give us, and some really high end IB connected storage.

The lead dev's requirements kept getting heavier and heavier despite blowing gobs on leading edge hardware.

Eventually we had an all-hands meeting where someone in sales did a mockup of our product in Excel.

They went to a customer in our target demo and loaded up the full dataset in Excel, and got the answers the customer was looking for. Sales had successfully implemented the product our customers wanted with no absurd HPC systems, but only Excel and a MacBook Pro.

I was reaching out to prospects before my first beer made it to me at the bar that night.

Still sounds like it was a fun lark for a while.
"The company is really just a consulting firm that deploys custom solutions for clients: this is totally fine by me, because it's really interesting and fun. But, the CEO is going to investors pitching the fact that we've built/are building some deep learning engine that can be applied to all problems."

Sounds like IBM Watson.

That’s pretty much all startups, sorry to tell you.