Technical Due Diligence and Venture Investing — How Much Is Enough?

When I was running reviews for tech magazines back in the nineties, a very well-known large computing company had a breakthrough they wanted to show us. Pitched as a powerful mainframe computer in a box the size of a PC, it had the potential to upend the computing world. Of course we wanted to see it, test it and play with it.

The company said we could only see it if we flew up to their labs. So we did. And yes, that system seemed to do all that was promised and more. At least until one of our skeptical reviewers noticed the big cable snaking out of the conference room we were in, and into the closet.

“What’s in here”, he asked, and before the horrified PR people could respond he flung the door open, revealing a technician sitting in front of a terminal, apparently feeding commands to a bigger computer, and then funneling the results to the screen on the conference room table.

These and other promises of technology’s utopian visions — many of which never materialized — created a streak of skepticism wider than a PC Jr’s keyboard inside of our reviews team. Whenever we were presented with a suitably advanced technology, we always looked for the magic pixie dust inside. And we often found it.

I’m constantly reminded of how new technologies can be overhyped — sometimes maliciously, but more often simply because tech entrepreneurs are as optimistic as I am skeptical — if not more. And now that I’m venture investing at Social Starts I see a similar situation unfolding. But unlike in our test labs, technology due diligence isn’t routinely implemented, particularly during early stage investing.

When you’re investing in apps or other consumer software products, that’s less of an issue, because typically there’s not some breakthrough technology that will enable the service. But with hardware and enterprise services it’s different.

In the year since I’ve been working at Social Starts I’ve evaluated two patents, tried to make sense of new video compression schemes, wrapped my head around enhancements to Bluetooth and WiFi, looked at whether a variety of VR companies are viable and tried to plumb the depths of natural language processing and mobile APIs. Although reasonably technical, I am not an expert in any of these technologies, and for the most part I’ve struggled to deliver a 100% perfect bill of health. Reviewing tech products is hard. Tech due diligence on early-stage startups where there’s often very little to poke at is even harder.

Luckily I have a lot of friends who know a lot more about key technologies than I do. And I’ve enlisted them occasionally when I need a deeper look. To date I’ve repaid them with drinks and dinner, but I’m loathe to go to back to that well too many times. I’m apt to start wearing out my welcome.

And this raises the broader issue of what level of due diligence is appropriate for venture investing. The topic has recently been in the news, as Theranos and UBeam have both been singled out for alleged shortcomings in their technology approach. This column is not meant to address either of those companies, nor the issues raised. I’m not an acoustical engineer, nor am I a biologist or doctor. And in the case of UBeam, two investors that I particularly admire — Mark Suster and Ben Parr — give me confidence that they conducted extensive due diligence before investing.

But the question remains — what type of tech due diligence should venture investors pursue? During our process at Social Starts we dive deeply into the business, extensively evaluate a CEO and CTO’s background and demeanor, and vigorously debate the company’s chances for success. And yes, when it makes sense during our seed investing process, we dive into the technology as well. At least for me, when I can’t understand a technology approach to my own satisfaction — or that of someone I trust deeply — I recommend we pass on the investment, even if it might otherwise be promising. The A Round, for us, is different. There we rely on funds that do dive deeply into their own tech due diligence, and who then share that with us.

That’s much more than many firms do. Many financially driven investment companies will dive deeply into the model, yet effectively ignore the tech. Often they will assume that one of the other co-investors has done the work — for example if a tier one Sand Hill Road firm is in the deal, then the tech HAS to be good, right? Not always. The lack of strong technical competence among many venture firms and angel investors also leads to a variety of relatively simple copycat companies being funded. This is why you see so many “Uber for food” delivery companies get funded. Few have the deep tech understanding to evaluate the next quantum computer or multi-nozzle 3D printer. But we all eat, and we all get Uber’s model.

At Social Starts we either build in-house expertise in an area we’re focusing on, or bring on a venture partner with deep domain knowledge. Our latest partner, Jon Gosier, for example has deep domain knowledge in the arcane worlds of big data, as we’re seeing more and more captivating opportunities in that space. Jun Li is a SAAS and cloud expert. Charles Smith has deep experience with marketplaces and media as well. And we don’t invest in what we don’t understand.

But how much due diligence is too much? Venture investing is all about taking a risky stance and hoping for a big outcome. 90% or more of many firm’s investments fail completely. But it’s also about de-risking that investment wherever possible. That’s why the industry tends to favor founders from the top schools, from the big four (Amazon, Google, Facebook and Apple) and with a prior startup success on their resume. But I think we as an industry need to go further. Whether you’re investing a few hundred thousand, or tens of millions, tech due diligence is absolutely essential when the products are complex. Otherwise you might just end up owning a piece of something that’ll soon end up in a closet.

What do you think? If you’re an investor, how much due diligence do you do? How much should we do? Leave me a note in the comments!