Quick Take — Avvir
Quick thoughts on a PropTech startup focused on deviation detection via digital twins.
This is the first post in a series I am starting to give readers a look at different startups in the Proptech space. These are meant to be quick looks under-the-hood at startups that come across my desk.
This “quick take” is meant as a first impression of the company and its competitive positioning. It is NOT a full underwriting (which can take months) or recommendation of investment. Just some candid first impressions of the company that are hopefully useful for its founder and for other investors.
So, here is a quick take on Avvir:
Niche: Deviation Detection via Digital Twin creation (more on that below)
CRE or Construction: Construction
Founder(s): Raffi Holzer
Concept — The basic idea here is that if you create a digital copy (twin) of a building from your architectural plans and then carefully measure the building as it is being built, you can detect errors in construction (“deviations”). MK — This is a huge trend in the industry. Creating digital twins to simply capture and contain (forget analyzing and acting on) the terabytes of data flowing in and out of commercial buildings is unequivocally the future. If Avvir can differentiate itself, this is a great niche to play in.
Problem Solved — Expensive change orders. The idea is that by the time you know that a steel beam is 6 inches off, it’s too late to fix and you have to redo the entire area of that floor. That can torch your budget and Avvir’s theory is that if you know about deviations from the plans sooner, you (the Project Manager or GC or Owner) can prevent that expensive change order. MK — I’m not a GC or PM, but I do know how devastating change orders can be for developers/owners and you wouldn’t offend me if you said that budget overruns are the main problem in modern construction.
Secret Sauce — Hardware agnostic. To get the as-built measurements, you need to have someone scan the building with some frequency (every day?). You can use any off-the-shelf scanner that uses Lidar and, soon, solid state scanners as well. MK — I like this idea on the surface because, like most VCs, I prefer the wonderful margins of software companies over hardware companies. But, when I dig deeper into the use case, I still see you having to teach construction workers how to use 3D scanners. I haven’t used them before and I’m told they are not THAT complicated, but you are still teaching a construction worker to use a new tech product. That will either slow your mass-market adoption or make your target costumer that 5% of project managers who are tech dorks like me. See Market comments below.
Competition — Plentiful. The splashiest of them is Doxel who made a sexy video (below). But they build hardware and Avvir is software only. Holobuilder is another one that is already well known and you could argue Matterport ($100M+ raised) is in this space. Plenty of capital already deployed in this niche if you consider those guys competitors.
MK — Overall, the track-what’s-actually-happening-on-site thesis is a well-worn path. Most people don’t debate the concept that owners of buildings do not have enough information about what is actually happening on their construction site until their GC comes to them with the bad news. Anything addressing that issue seems to get meaningful consideration. But, that also means that everyone who has ever worked on a construction project and then ventured into tech is trying to solve that problem. So, while Avvir may not have any direct competitors in the hardware-agnostic deviation-detection space, I’m 98% sure that landlords are being approached with a dozen other concepts claiming to be “the data you need to actually know what is going on with your new building.” How do they stand out as the one tool to rule them all?
Addressable market — Avvir claims these are mission-critical buildings (data centers, healthcare, municipal). I like the property types they are addressing because they will be much less prone to market swings and constrained construction budgets in a recession. MK — The great thing about the Proptech niche is that the addressable market is almost always huge. It’s always in the billions of dollars because of the shear size of the construction and CRE industries. When I first saw Avvir a couple years ago, my concern was that the number of GCs/Owners who would pay for this is only the top 5% who are working on the largest projects in the world. That is a very small addressable market. If this can be an affordable solution with a great UX, then it may have the legs to survive a recession on the back of these somewhat-market-agnostic properties.
Team — Raffi. Tira. Cody. And Cory. Looks like about 8-ish employees according to LinkedIn. MK — All in all, a decent team. No exits that I can see, but everyone seems to have some relevant experience in the industry (Bentley Systems and FieldLens, for example). The tech team, Tira and Cody, seems like they are young but have some pedigree (Amazon) on their track record. I see Raffi balanced some of the inexperience with having some advisors with (sometimes metaphorical) gray hair like Tracy at PlanGrid and Dan K at Leopardo Construction. With the advisors, I would call this a B+ team. Not market-leader or breath-taking for the stage of the company, but still well above average and not a clear weak spot.
Pricing and Revenue — The use case they quoted to me is a $75M building and they project their revenue totaling just 0.04% of cost. MK — As a % of project cost, it’s almost nothing. Sure, but it’s also probably an additional cost on a construction project with already thin margins (margins that have been further shrinking for a few years, by the way). My hunch is that you aren’t selling this to GCs as a tool to use versus other tools. I’m betting you have to convince them to use this at all instead of using nothing to track deviations. Perhaps that’s the reason for the focus on the owners and developers rather than the GCs? If the lady paying for the building says you have to use Avvir to track progress, then it’s just part of the budget and your bid.
General Thoughts and Further Diligence:
My problem with this company is that I like Raffi and I need to be careful not to let my appreciation for him as an entrepreneur cloud my views on the merits of the company (know your biases!). Solving change orders is a huge problem in construction, but, from what I understand, is also where most GCs actually make their margins this late in the competitive building cycle. So I assume GCs might be a little hesitant to give that up. The team is good for their stage and their tech seems relatively straightforward (reconciling CAD and BIM models with real-time point clouds). As a fund manager with Seed stage allocations, I would do more diligence with my network to see if this merited a term sheet.
What I would need to dig into —
Do GCs want to use a tool that lowers change orders? If not, what is the developer/owner appetite for this?
How long does it take to train construction workers on the scanning technology necessary to populate this?
Do any of the competitors I find have a directly competitive offering or are they likely to pivot into this niche given how much capital has already been deployed here?
I need to dig into the margins and see how those trend at scale.
Those are my quick thoughts on Avvir. If you want more info, reach out to Raffi on LinkedIn here.
Disclaimer — This overview is not meant as a thorough representation of the risks of investing in a startup and should not be constructed as such. Please consult your financial advisor about the appropriateness of investing in early stage startups in the Proptech space as a part of your financial portfolio.