URBAN-X Startup Selection Process

Tristan Bel
Published in
5 min readApr 4, 2023

Following “On Improving the Accelerator Model”, this article covers a critical piece of running an accelerator: the selection of the founders. After completing 12 cohorts with a few hundred applications each, how do we find the best 5–9 candidates? And how do we do that efficiently?

Using a CRM

If you’re receiving more than a 100 applications per cohort, and have more than one person reviewing them, you’ll want to use a CRM. We use HubSpot but there are many comparable tools available, some specifically designed for accelerators.

Using a CRM facilitates collaboration, ensures transparency and traceability, and makes the process more efficient. Every application, no matter how bad, deserves a response and automating these responses saves a lot of time. Once we decide to interview the founders, we usually move off the CRM and use email and google docs.


Sorting out applications that are very low quality or off thesis goes fast. We keep the online application form short. There are some applicants that do not provide sufficient information to make an initial judgment, and if there’s enough to pique our interest, we’ll follow up with questions. We know that to many founders, there are many accelerator programs they can apply to and we don’t take an initial low effort personally. We won’t ask twice though.

We help startups that are not a good fit sort themselves out before they click “submit.” On the application page, we provide a link to an article explaining what we’re looking for in detail and re-emphasize our urbantech focus. That acceptance rate doesn’t look as good, but this makes the application review more efficient.

Once we have the information we need, we spend more time doing general research. Most of the time, we have a lot of background information already since we already have a few startups in our portfolio in specific sectors — building decarbonization, EV infrastructure, circular economies, community engagement… these are topics we’re familiar with. Sometimes, we have a pre-interview call for a quick exchange about what stands out to us and also to get a sense of the founders.


From the information we gather from our prior knowledge/research/pre-interview call, we come up with a list of more thorough questions which come in 3 or 4 sets depending on the target market and whether or not the product has a hardware component.

This first interview is led by our investment committee and runs for 45–60 minutes. We search for answers to our questions and also look for team dynamics and coachability. How are the founders complementing each other, do they “pass the ball” seamlessly, and most importantly — are they open and interested in learning. Do they know that they don’t know everything? This is important to us because if they’re not ready to at least consider our input, we’d rather not waste anyone’s time. And if they’re not curious, they probably won’t be doing genuine customer discovery which is a cardinal sin in the early-stage startup world.

We have a firm rule about advisors participating in interviews: they’re not allowed to. We made that decision after a number of founders delegated too many responses to their advisors. It’s a problem when some of the core knowledge and competencies are outsourced to individuals that are undoubtedly highly qualified but unfortunately not invested enough in the startup. We want to know that the founders have integrated what they need to know from their advisors and are not using them to make their points.

Data Room

If everything checks out, we request access to the startup’s data room, covering market size and financials, cap table, product and IP documentation, and basic legal information. We don’t ask for extra work — just that they share whatever they have — the timeline doesn’t allow for much anyway. This also allows us to judge how ready they are for investors, and how organized and thorough they are in general.

Interview with EIRs and Scoring

From the first interview and with the data room documentation, we’re able to drill down on the product, market assumptions, and business model. We bring in two of our experts in residence and match them with a startup depending on its characteristics.

We then use a scoring sheet to rate:

  • Product/Solution: How unique and new is it? Does it solve a real pain point? What is the solution’s potential climate, social and urban impact?
  • Market Size and Access: Is there a large, global market? Is the idea scalable and has the team identified a good beachhead market?
  • Team: Is the team in love with the problem? Do they have unique, deep domain knowledge? Do they have founder experience? Do they complement each other’s background?
  • Distribution: Has the team identified the most efficient channels to scale their commercialization model into a sustainable and repeatable business?
  • Competition: Does the team have a good understanding of the competitive landscape? Do they differentiate well against competitors? Do they own IP, or have unique defensible advantages over their competitors?

The score gives us a sense of ranking, but to make our final decision, we also look at how much we’d be able to help them with our resources, and that we give support to otherwise underrepresented founders.

The Memo

Once we’ve made our decision, we draft a memo with the founders. The first part of the memo summarizes our understanding of where the startup is at following the 5 categories above, as well as their score. This serves as a common ground of understanding between the startup and our team. Secondly, the memo outlines the work we’re planning to do together and the overall outcomes we’re looking to achieve.

Once we’re in agreement, that part is pasted into an offer letter or program agreement that also contains boilerplate language and a code of conduct for accelerator participants.

At this point, we’re ready to start and we schedule the onboarding meeting with the entire URBAN-X and founder team and to kick off the accelerator program.