Evaluating Startups
Recently I’ve been pondering the ways startups get evaluated… particularly the tech DD side of things. This all goes out of the window, as there are often factors far far more important to investors than if the technology works1. But hey, it’s what I’m interested in.
Phase 1: Basic Technological Feasibility
Does the basic technology have sufficient technological support based on public information, academic publications.
For example, do multiple scientific publications from different groups support the general approach. For example, let’s take sensing of a novel Biomarker with a Nanopore.
Can you find multiple groups using similar nanopores in the literature?
Can you find multiple groups detecting similar sized/charged particles using these nanopores?
If yes, great. If the only supporting work comes from a group associated with the startup… maybe not so great.
Phase 2: Internal Data Evaluation
The startup may have generated some data. Evaluate it in light of the scientific literature discovered above.
For the Nanopore example above, do the translocation rates and deflection levels match the literature? Is there some reason they wouldn’t match?
How comprehensible is the analysis presented? Do they walk through the raw data to the final results? Or are there a few squiggles and a black box labelled “machine learning”?
If the approach relies heavily on machine learning are you confident that experiments were well controlled/randomized?
Phase 3 - External Validation
Something, anything, that shows the group has built something related to the company thesis that has been evaluated by a trusted external partner.
This moves away from technical DD and gives an investor much more confidence that what the company is pitching is actually real. Startups make ambitious statements, and it’s generally in their interests to do so.
But if you’ve got an external party, so say “yes we ran some samples, and it looks real”. This is a huge leap forward from “we have some internal data”.
This probably helps even more if your start is a relatively unknown entity.
The important thing here is that external validation doesn’t need to be on the complete system. It can be one component, or the basic detection approach, something for which there isn’t a viable market is fine.
It’s most about saying “yes, these folks can build things and this part of the basic technological approach can do something which supports their basic thesis”.
Phase 4 - Early Access
By this point you’re pretty much done. Early access partners are providing feedback on the technology and you know that the technology works.
One caveat here is that early access partners are often happy with anything that works and that will let them publish a novel paper with a new technology. It’s therefore important to evaluate the platform with respect to the market. I.e. is this just something orthogonal to what we have already? but it doesn’t provide anything new? Is it so expensive that the market size is tiny? etc. etc.
Summary
I rarely follow this exact process when looking at startups. But I think I often get pushed into using one or more of these approaches. In particular if the data presented by a startup makes me feel “uneasy” I’ll often revert to “Phase 1” approaches and try and get a general understanding of how much of a “technological leap” over the state of the art is being proposed.
Sometimes, that’s too much of a “leap” for me to be comfortable with…
Investors are in general looking to how they will get a return on their investment. Strictly speaking this occurs in exactly one way, another person buys their investment. This can be through a number of mechanisms including:
Acquisition
IPO (someone buys their shares on the public market).
Another investors buys their shares.
But none of these are “the company is profitable and returns dividends”. And most investors will likely be hoping to get a return before the company is profitable.
As such the question often becomes something less like “is this company likely to be profitable” and more like “can I convince someone else that there is a reasonable probability that this company may be profitable, or have something that another company needs to acquire, or other offset some other investment someone else has.”
That then gets a lot more complex and incorporates factors such as: Regulatory risk, market risk, who the founder plays golf/football with etc. etc. etc.
This may in part be why investment sometimes look irrational from a purely technical stand point.