How to assess your portfolio’s risks of ‘unintended consequences’
Early 2019, we became (to our knowledge) the first VC globally to map and share the risks of our portfolio’s ‘unintended consequences’. For those not familiar with the concept, unintended consequences are outcomes of an action that are not intended or foreseen. They can be positive or negative, but it’s most likely the latter that pose risks to a company and its investors, filling the news with headline grabbing stories. From the misuse of personal data to increased screen time to social media-triggered anxiety – these are all examples of unintended consequences of many tech startups (naming no names).
Despite pursuing positive outcomes, tech for good is no less immune. As investors who centre our investment thesis on delivering positive social or environmental impact, we have both a moral and financial incentive to ensure that our ventures deliver impact as intended.
At BGV, we started thinking about assessing our portfolio’s unintended consequences as early as 2019 when we first asked our portfolio to tell us about what negative effects might arise from their product or service. This year we built on those efforts by grouping those consequences and mitigation strategies into distinct categories and sharing them for the first time in April 2021 in our latest Impact Report.
With an interested response from many in the wider tech for good, VC and impact investing ecosystems, we thought we would share how we assess our portfolio’s risks of unintended consequences with a guide on how other investors can too.
Why do it?
Despite containing the words ‘consequences’ and ‘risks’, this analysis should not alert the compliance team and is in fact, an opportunity to maximise the positive impact of your portfolio. Eliminating risks altogether will never be possible but knowing what those risks are can help to understand the relative importance of each to the business.
This guide is intended to help investors assess unintended consequences across an entire portfolio. By understanding your portfolio in this way, you can not only help mitigate the risks but also support your ventures in building better products and services.
The insights from this assessment should also provide valuable information that can guide the support you provide to your portfolio and ultimately help founders deliver on their impact outcomes.
1. Assess your aims and select a framework
It’s useful to start by thinking about what you want to understand. When we started thinking about how to do this exercise, we wanted to categorise impact risks by type and see which ones were the most relevant to our portfolio.
In our preparations, we reviewed three frameworks that could help to analyse unintended consequences: Doteveryone’s Consequence Scanning framework, Omidyar Network’s Ethical Explorer Toolkit and the Impact Management Project (‘The IMP’), are all brilliant and offer differing ways of approaching this task.
To give some context, the Ethical Explorer Toolkit hones in on risks most commonly associated with tech products, such as surveillance, algorithmic bias, online harassment and market inequalities. It is great for laying the foundations of ethical product building. However, in our case, we found that as many of our ventures are still pre-product or early-stage, these risks were not the ones that they were most worried about. For us, only covering tech risks would not have been representative of our portfolio.
Doteveryone’s framework offers a clear structure for initiating a discussion about potential consequences – we would recommend using it if you need more buy-in for engaging in this exercise in the first place. The framework offers some categories of risks but once again, several were focussed on digital technologies. We felt our portfolio’s risks were broader and that for analysis purposes, a different framework might work better.
We went with the IMP because the impact risks reported by our portfolio mapped more easily onto their taxonomy. For many of our ventures, the risk that a prospective partner might dropout from using their service or that a user will misunderstand what their product is about (stakeholder participation risk), was much more relevant than any tech-specific risks that are outlined by the Omidyar’s framework, for example. In essence, the concerns of early-stage companies were more focussed on finding their place in the market, something that we felt was more directly addressed by the IMP.
2. Ask the questions
To streamline the logistics behind this process, we asked our portfolio questions on unintended consequences in our quarterly reporting form that they’re used to completing.
We added two additional questions:
- a. What negative effects (unintended consequences) might result from your product or service?
- i. What are you doing to mitigate them?
- b. Did your users have any bad experiences with your product or service? If so, what were they?
Answers to question (a) will vary but you should be able to see fairly quickly which ventures are aware of their unintended consequences and which are yet to think about them. Answers from companies that do assess their risks will be your raw data that you then categorise according to your chosen taxonomy. It will also allow you to see who is already mitigating any possible negative effects and who might need some help.
Question (b) on the other hand gives an insight into whether any of those unintended consequences actually happened in real life.
3. Categorise and add up
In categorising the responses, we went through every single one and marked the type of risk it represents, how the company might be mitigating it, and the likelihood of that risk occurring. Some risks are easier to categorise than others and in the end, you might have to use your best judgement to make the call. Top tip: be consistent. It’s helpful for one person to do the exercise and for another to review for consistency in categorisation.
This exercise did not result in a beautiful spreadsheet but you can see the general structure below:
Once you fill this out, you will end up with a categorisation of your portfolio’s risks. You can see a summary of our portfolio-wide risks in the image below but there’s nothing stopping you from experimenting with this further. You can repeat this exercise to see what your risk breakdown looks like for different funds, investment themes or company stages to see if various segments produce different trends or require different approaches in support.
For example, by doing further segmentation, you might find that later-stage ventures are more concerned about ‘efficiency risks’, such as delivering their outcomes using too many people or at too high a cost. For earlier-stage companies, the risk that their staff incorrectly onboards a new user (execution risk), might be more prevalent as people are still getting to grips with a new product. Cutting this data in different ways could help to get a more nuanced understanding of risks and how to address them.
4. Learn from your results and plan actions
This data will become much more meaningful if you use it to reflect on your role as an investor and the support you can provide with these new insights.
At BGV, for example, we have identified stakeholder participation risk as a common risk and one that is particularly relevant for our early-stage ventures. This one relates to the probability that the expectations or experiences of stakeholders are misunderstood or not taken into account. CBInsights research actually shows that this is one of the top reasons for why startups fail. We try to help our ventures address this by communicating the importance of a user-centric, iterative approach in building the product. We do this throughout our engagement with the startup – assessing it as part of our interview process before we make an investment and by running a two week sprint on ‘Designing with people in mind’ to give just a few examples (we promise, we did not rename it for this article).
This exercise was also valuable to us in pointing out what risks our ventures do not perceive to be material at their stage. Alignment risk – the risk that impact is not locked into the enterprise model – comes at only 3%. This might be because mitigation strategies for alignment risk are inherent to how BGV filters investments and conducts its investment process. For example, we only invest in ventures where direct impact is intrinsic to their business model. We also stipulate that companies must adopt a purpose clause in their Articles of Association in order to receive investment, making sure that the mission is ingrained in their business even as the company scales or changes leadership.
BGV’s results as an example
In our most recent survey, we found that 78% of our portfolio respondents actively assess potential unintended consequences arising from their products, of whom 77% have identified ways in which they could mitigate them. The more of your ventures that are analysing their risks and implementing actions to mitigate these risks, the lower the risk of your portfolio and consequently, the higher the chance of returns.
Our target for next year is for 85% of our portfolio to actively assess unintended consequences with all ventures who have identified risks to have also given us ways of mitigation. To achieve this, we will continue to encourage our ventures to think about their risks early on through our programme and will offer tailored support to help our ventures address their specific concerns.
As investors, balancing risk and reward is at the core of what we do. We hope that this ‘how to’ makes it easier to apply that thinking not just to financial returns, but also impact.
This guide was written by BGV’s Operations & Insights Manager Yumi Tsoy