Formadness

How do you build a ‘tool’ that replicates an iterative process? That has been the challenge on a project in collaboration with Transparency International and a development finance institution. We’re creating something to help impact-focused private equity funds identify and manage risks in their portfolio companies (PortCos). It’s much more complicated than I anticipated.

If due diligence and risk assessment are working, they’re adaptive processes. Below, I’ve tried to explain the classic flow of our everyday impact investment projects. The green cells are mainly for us to do; the purple ones are where we prevail on the subject of the DD (in this case, a PortCo).

Putting this into a ‘tool’ (Excel, alas, as most people want a format they can download into their domain instead of an online version, which could build in logic) is chaotic.

Many DD forms and processes are gruelling because they’re trying to fit into an Excel logic sequence. They might go:

🚦 Fill out a form with corporate information

🚦 That is then used to run the names through watchlists

🚦 Fill out questionnaires (have you ever been… and do you have…)

🚦 Someone reviews this and asks for supporting evidence

🚦 Submit evidence for the “do you have…” questions

🚦 Someone reviews those documents (best case) and may follow up

🚦 Deal with the follow-up

🚦 Review the reams of information (maybe use a generalised ‘score’)

🚦 Go or no go decision

🚦 Pick from a few different contract terms (sometimes) to reflect risk

So what’s different between the Excel and the image above? Mainly it’s the first step. We look at the nature of the proposed transaction and relationship. A $10m loan to a 50-person hydroponics company is a very different DD process than a $200m investment in a renewable energy company spanning four countries. This is intuitive, of course. But Excel isn’t.

So, the question to ask (of your DD and risk assessment process): Was it built using Excel thinking or human logic? If the former, don’t worry. The first step in the image can be made quick if you compile a dataset of questions (and documents to request) tagged by sector, size, and nature of the proposed relationship. When we start to ask the hydroponics company questions about controls relevant to a large energy firm, we’ll be wasting a LOT more than the 20 minutes (or so) it takes to get that first step right.

Tangentially, we’re at a very early stage of discussions around ways to build interactive tools (potentially leveraging AI) to make the green cells quicker (the 80/20 principle—80% of the heavy lifting in 20%, or less, of the time). If that’s something you’re also considering, we should talk.