When I started investing in early stage edtech I didn’t have a specific model, financial or otherwise, to judge who received my small investments.
I have spoken and written about what I subsequently developed, although my models and their underpinning are neither scientific, mathematical or grounded in any accepted business thinking.
I started out by simply talking at length with founders, almost all of whom were young and inexperienced. One of my additional ‘tricks’ was to meet where we would have to use the escalators at a tube station. My thinking here was that those who stood on the right were subliminally employees and those who walked on the left had some sort of unconscious urge which I felt was necessary in any entrepreneur. Later it was pointed out to me this was rather discriminatory (hat tip to Jodie Lope) as some people have disabilities that prevent them walking up an escalator. Next (2014) came Taylor’s 10, a quick quiz (2021 version here) to test what the sitter knew about the wider history of the business of edtech. I wanted knowledge that cannot be easily Googled, indicating a deeper interest in edtech, beyond the limited window of what their startup was claiming to fix. This tangentially linked to my preference for independent thought rather than the Stupidity of Crowds.
Eventually, I distilled down into a simple acronym what I thought were the three key issues of investability (applicable generally to all businesses whether startups or further along). This was POP:
- Product Market Fit
- Operational Capacity
- Profit.
However, over the last three years as my interest in educational research has grown, especially around what constitutes evidence, I started adding to POP. First was POPE (evidence) and most recently POPEYE.
- Product Market Fit
- Operational Capacity
- Profit
- Evidence
- Youth
- Experience.
The last two are probably a subset of Operational Capacity, in that I think most successful edtech businesses need a mixture of the enthusiasm of youth allied with people who have had more experience in running successful enterprises. Several years ago, at a talk about discrimination in the sector, I made myself unpopular by arguing that the single worst and most pervasive type in tech and edtech was ageism. I was the only speaker who done some primary source research about this, having interviewed 25 edtech companies I knew (some have since gone broke) about factors such as the percentages of their staff who had tertiary degrees, came from outside the UK, gender(s) and age. On most metrics these edtech companies were on a par with UK plc as a whole, except for employees aged over 40. The two companies with employees in this age category were also the standouts in terms of POP. Cue a lot of hand-wringing before the host proclaimed, ‘The bar is open’, and the issue evaporated almost as swiftly as the alcohol (which I had foolishly sponsored).
Evidence came later, the result of following the activities of researchED and the crystallising of twenty years’ reading of not just pitch decks, but also hundreds of research articles. These ranged from one about the impact of tech at Westpoint Military Academy (negative) to one from the National Literacy Trust (almost 20 years ago) about the impact on oracy of forward-facing prams (negative). As my evidence journey continued, I learned that not all research and evidence is equal and to always dig in and look at not just the sources, model and numbers, but also at wider issues like the speed of projects (I was very persuaded by the rapid models developed by Professor Rowland Fryer) as well as replicability, a fundamental weakness in all social science research and especially in education (shout out to Pedro De Bruyckere). Edtech Evidence, founded by Dan Sandu, is another newer group who are pushing the evidence agenda forward.
The problem with evidence is that the more you know the less you know or at least the less you tend to trust. A recent example is:
- A New National Purpose: Innovation Can Power the Future of Britain from the Tony Blair Institute for Global Change and Lord Hauge. On page 33 of this document it claims: ‘One AI-powered teaching and learning platform, from UK startup CENTURY (CenturyTech), has already been found to cut workloads by an average of six hours a week, freeing up time for teachers and social workers to focus their attention on supporting children while spending less on paperwork.’ Impressive, except that the report’s cited source is a web page promoting UNESCO’s 2019 Mobile Learning Week, where the founder of CenturyTech made this and other claims. To my knowledge, CenturyTech has never provided any evidence to back up this significant claim (I’m happy to be proved wrong and to see any such data). What this shows is that if a credible source publishes something and it’s repeated via a reputable organisation (in this case UNESCO), it is taken as a fact when it may more precisely be an urban myth. The BBC’s More or Less programme unpicks similar myths such as, in 2017, the claim that 65% of jobs that children now in school will do have not been created. This statistic turned out to have ‘jumped the shark’ after Cathy N Davidson’s 2011 book, Now You See It, where she claimed this came from a report published by the Australian Innovation Council. Yet according to the Australian government, the report and Innovation Council have never existed. Tracking back to the recent Blair Foundation document, it would be terrific news if a UK (or any) edtech company had solid evidence that their AI product, when used in schools, had saved teachers six hours per week. Sadly, this claim, like the one dissected by More or Less, does not (at this time) stand up to scrutiny. After 30 years in the business of education and over 20 in edtech, my unscientific tests are just that. While they may have evolved into POPEYE, smart investors and anyone interested in education should get involved with and read more research. Perhaps they should take a lesson from the alluvial gold miners, who in the 1850s flocked to pan for gold in the tiny Australian town where I grew up. Most miners spent their time panning for tiny glints of gold amongst the rock and sand, an activity akin to finding credible evidence in edtech. The lesson here is that the people who got richest weren’t the miners but the people who sold them the picks, shovels and pans.
