Why I Hate Your Market Size Slide
A version of this article was previously published on Propmodo in Sept 2018
Given how much topic annoys me, it’s surprising that I haven’t written about it before.
As managing director of Dreamit UrbanTech, I see well over a thousand (if not 2000) pitch decks a year. Now that we work with more mature, pre-Series A round startups, most of our applicants have already raised a seed round and may have even gone through other accelerators, so it’s more than a little bit surprising to me that somewhere between 10-20% of those applicants still have unnecessarily complicated Market Size slides.
Here’s an example of what I mean:
Yeah, you’ve seen it before too. Looks kind of impressive, doesn’t it?
Unfortunately, here’s what I see:
Let me break this down for you you:
- Total US Construction – The startup sells software. Why do I care about money spent on concrete, labor, etc.? That’s no more relevant to this startup than Dim Sum sales in Moscow.
- Contractor Software Spend – They sell software to subcontractors. If a general contractor is spending $50K or $100K a year on Procore, why does that matter? Those GCs have no use for this startup’s product, will never use it, and will certainly not be spending any of their money on it. Might as well toss in revenue from video game sales for all that has to do with this market.
- Subcontractor Software Spend – Right type of product, right customers – now we’re getting somewhere. Probably still a top-down estimate but at least this is less obviously wrong.
Top down is for convertibles, not pitch decks
Let me let let you in on a secret: top-down market size estimates are almost always incorrect and generally useless to an investor.
Why are they almost always incorrect? Because they typically capture a lot of related spending on different types of tools. In the example above, the figure includes spending on other construction related software, not to mention more general purpose software like Microsoft Office and Quickbooks. Furthermore, the top-down approach ignores pricing. Let’s say, hypothetically, that you make software that replaces absolutely everything the subcontractor might possibly buy but do it at a tenth of the price. By definition, your market is 90% smaller than the top down estimate suggests. Alternatively, say you provide a solution that’s much more functional and valuable than anything on the market and expect to price at a premium to existing software. In that case, the top down market size estimate understates your potential.
The more accurate and useful way to estimate your market is bottoms-up. It’s really not that complicated. At the end of the day, it is simple third-grade math: Total number of potential customers times what you plan to charge. Really, it’s that simple.
So that’s what an appendix is for…
I called top-down estimates “generally useless” instead of simply “useless” for a reason. They occasionally are a useful appendix slide as a sanity check. If you are creating a completely new product category and your bottoms-up market size estimates are large, it’s good to know the size of the total budget you are competing for. For instance, if your bottoms-up estimate comes to $100B and the software spend for the entire construction industry is $130B, you are essentially arguing that your customers will either stop buying 77% of all the software they currently use to buy yours or that they will manage to steal budget from other departments. Not entirely impossible, but extraordinary claims require extraordinary proofs. Needless to say, you’ll get a lot less pushback if your bottoms-up estimate came to ‘just’ $10B out of $130B.
Who’s on first?
As simple as the above TAM equation is, it still has two variables: number of customers and price. So if you haven’t told me what you charge, the equation isn’t going to make much sense. You need to define the price variable before adding the number of customers into the mix. That’s why your Revenue Model slide should almost always immediately precede the Market Size slide.
In some cases your Revenue Model may be a bit complicated. Your price may increase based on usage (e.g., number of seats or concurrent users), frequency (e.g., API calls, reports or searches per month), or value received (e.g., which modules they subscribe to). Your Revenue Model slide should include this detail at a high level but should also clearly show what you believe the average spend will be across all your customers. Assuming the reader thinks that estimate is plausible, he or she can seamlessly plug it into the market size equation on the next slide.
Maybe we can meet in the middle?
Typically, when you see the infamous 3 circles the only number that matters is the bottom circle but, alas, not always. Sometimes the Total Addressable Market is actually the middle circle. For example:
Now we have to actually think about it (dammit!). If the startup is pitching software for roofing subcontractors, the bottom circle is the relevant one (and I can stop reading right there because that market size is way way way too small). But if a founder is making software for all subcontractors and they’ve identified roofing subs as the first subsegment they are targeting as they go-to-market, then the middle circle is the one that matters when it comes to TAM.
Bear in mind that the bottom circle is not wrong. It’s even interesting and relevant information as far as an investor is concerned. It is simply in the wrong place. “TAM” means “Total Addressable Market”: if everybody who could use your software uses it, how big is the market? If you choose to sell to some types of customers before others, that information belongs on your Go To Market slide.
Don’t sell yourself short
What would you make of this slide?
At first glance, this looks like a $200M market. That’s a nice size for a self-funded company, but it is generally not considered large enough for venture funding. But take a look at footnote 3. This company is assuming that they ultimately get only 25% of the total market. This may very well be true but they are selling themselves short. Investors think in terms of total market size; all our rules of thumb implicitly assume that the startup will only capture some of that total. In theory, we should be able to correct for that but, psychologically speaking, the low number sticks in our mind. Plus that assumes we caught this in the first place. As I mentioned before, we don’t particularly like it when you make us think.
When top-down and bottom up are the same
Before I get more than the normal volume of hate mail and ‘gotcha’ email, there is one legit exception to this rule. Under certain revenue models, the top-down estimate is actually what drives your bottom up number. For instance, if this software was used for purchasing construction material and had a revenue model where the software was free but they got a commission on all sales of, for instance, 4% of contract value then the top down figure of everything spent on construction supplies is actually the entirely relevant input to the “third-grade math” equation for market size: Total purchases x 4% = TAM.
You would still have to be very careful to exclude the types of construction supplies that are not on their platform (e.g., if you can’t sell cement, you have to pull that spend out) but in this case, the top-down data is not only valid but necessary.
“The TAM Commandments”
With the above in mind (and apologies for the awful pun), here’s a quick recap of what you can do to make your Market Size slide as effective as possible:
- Thy TAM shall be a single figure.
- Thou shalt have no other market figures before (or after) the TAM and expect Me to figure out which one (or two) to ignore.
- Thou shalt use bottoms-up estimates. Top-down estimates are an abomination. (Except when they aren’t)
- Remember thy Revenue Model slide and keep it before your Market Size slide
- Honor thy Go To Market with its own slide. Leave GTM strategy off the Market Size slide.
- Thou shalt not kill your Market Size by reducing by your expected market penetration
- Thou shalt email me if you can figure out a way to riff on adultery, theft, bearing false witness, and/or coveting your neighbor’s wife in this context.
Thanks for reading and please share this with all your friends… because I’m sick and tired of bad Market Size slides.
Note: This example is loosely based on Dreamit UrbanTech alumni Knowify. Their Market Size slide looked nothing like this and, for those of you who actually read the footnotes on the charts, the $800M TAM is 800K subcontractors in the US x an average software subscription price of $1000 per sub. So in other words, Knowify did it right. 🙂
PropTech Pitches That Are Past Their Expiration Date
A version of this article was previously
published on CREtech in Sept 2018
Coming off another successful recruiting for our 3rd Dreamit UrbanTech cohort, we had the pleasure to meet quite a few truly incredible startups.
This piece is not about those startups.
This is about the other ones, the startups that, like milk past its expiration date in a coworking space refrigerator, we’d really like to quietly disappear and be replaced with something fresher. So, after canvassing a few of my colleagues, I’ve compiled this list of startup pitches that, absent extenuating circumstances, we’d just as soon not see again.
It’s a community portal for tenants
I live in Manhattan. I don’t even want to talk to my neighbors in the elevator so why would I want this? In virtually all the buildings I’ve lived in, there has invariably been “that guy” (or woman) who has tried to rally the other tenants to be more social. Often, we like “that guy” a lot – he’s nice, he takes our mail in, signs for our packages, etc. We just have no interest in what he’s trying to do.
Kidding aside, there’s nothing about this idea that couldn’t have been done as far back as the late 90s which should be a huge red flag to any entrepreneur considering a startup like this. With so many hungry and talented entrepreneurs out there, good ideas don’t just sit around waiting. In fact, established companies like BuildingLink have community sections that are invariably ghost towns. If you have a burning conviction that the world needs a tenant community portal, you should consider the possibility that you are “that guy.”
It’s a real estate crowdfunding site… but with blockchain!
The most charitable thing I can say about these pitches is that they (or most of them, at least) were not ICOs.
We started seeing pitches for real estate crowdfunding sites as far back as 2014, if not earlier, and there are already a number of players in the space with significant head starts (RealtyShares, Fundrise, RealtyMogul, Patch of Land, etc.) so if you are a pre Series A startup in the space, you are pretty late to this party. Since these are basically marketplace plays, first-mover matters.
But wait!” you say, “we use blockchain!
So what? It’s not that hard to keep track of fractional shares in a building using an old-school, centralized ledger. If you standardize the legal documents and purchase process, you’ve already removed the friction on this process. The hard part here isn’t transactional friction but marketplace liquidity: you need enough buyers on the platform so that when someone wants to sell their shares (or tokens) in a property, there is someone willing to buy it. If you don’t have a deep pool of potential buyers, you end up with an asset like small cap stocks: easy enough to buy but hard to sell (especially in a down market!)
The possible exception to this rule are companies like Harbour who focus on tokenizing high-end trophy properties. These are the blue-chip stocks of the real estate world. There will likely always be smaller investors willing to own a piece of the Empire State Building. So giving its owner the ability to sell part of it to a mass market rather than to the current small circle of big players who can afford to invest at that scale both greatly increases marketplace liquidity and reduces transactional friction, unlocking (at least in theory) significant value for the building owner.
It’s a lead gen site for commercial real estate
I have the utmost respect for lead gen and, given the size of these transactions, there is potentially a lot of money to be made selling leads to landlords and their brokers. The trick is getting the tenant to start their search on your site… and you need to do it in a way that your competitors cannot immediately copy or else your cost of customer acquisition will be bid up until your margin is gone. Put another way, if you are using Google AdWords to drive traffic to your site, so can your competitors.
Zillow, for instance, succeeded in creating a site that residential buyers know to go to at the very startup of their home or apartment search by aggregating and cleaning up messy, fragmented public data and presenting it to the public in an easy to use interface. In theory, anyone could have done this but they moved first and fast, creating brand equity that’s hard for a potential competitor to displace without either creating something a quantum level better or spending a lot of money on advertising to launch a competing brand.
Our app helps community residents get in touch with their representative and get more active in local politics
If they wanted to do that, wouldn’t they start by at least voting? This is an example of civic tech backwards think: instead of creating an app to fill demand, they want to create demand with their app. And since here too, people have been banging their heads against this wall for nearly two decades, if you still think the public is just dying for an app like this, it’s very possible that you are “that (other) guy.”
We are a chatbot for residential brokers
It is telling that these startups rarely include successful real estate agents on the founding team. Converting a productive buyer into a client is mission critical, especially in an industry with little competitive differentiation. Agents convert products with personalized service and emotional rapport. A chatbot is the exact opposite of this and, as a result, agents are extremely reluctant to rely on them for this stage in the conversion funnel.
The rental side of real estate, especially on the lower end of the market, can be a brutal, time-consuming slog. Most agents transition from representing renters to other parts of the market as soon as they possibly can, leaving this segment to newbie agents or high volume / low service shops so it’s conceivable that a chatbot for renters’ agents might have legs….
We make 3D models from 2D floor plans
There’s value here if you can pull it off but there’s just not enough data in a 2D model to get to something buyer-ready automatically. So either the landlord has to customize the raw results a lot (too much effort for them) or the startup does a lot of post production (and becomes a service industry selling man hours rather than a scalable tech startup).
While not full-fledged startups, these phrases were often enough to make us gag all by themselves
Blah blah blah… drones!
Yes, drones are pretty cool and they do have the potential to change a lot of things, both in construction and real estate and the world in general. But if your startup is basically a glorified drone piloting service, you are selling man hours (not a model that VCs like to back), have no competitive advantage and no barrier to entry. To us, you are basically a taxicab company.
Blah blah blah… AI
So what exactly makes it AI (or Machine Learning for that matter) as opposed, for instance, to a simple database query? As the famous quote goes, “I do not think that word means what you think it means.“
… and the user gets a dashboard…
My car has one dashboard. Why would you expect a property manager to want 6 or 7? I’ve head the phrase “dashboard fatigue” a lot lately…
Instead of covering them another dashboard, integrate with their existing dashboard or, better yet, automate the responses to the data you collect so they don’t have to check a dashboard at all… or even think about it. Just. Make. It. Happen.
We’re Houzz meets Uber meets Robinhood
Here’s a hint: the Hollywood style analogy should get you an instant “Ah, I get it.” If the investor has to think about it to understand what you mean, it’s a #fail. I don’t care how cool you think it sounds, skip it and cut to simple description.
Acknowledgments: I’d like to thank Aaron Block, John Gilbert, May Samali… and all my other less brave colleagues who opted to contribute to this piece anonymously 🙂