data and other dangerous things
why this exists
Hi, I’m Osetohame or Matt, depending on which of my eras you met me.
I work with data, but that’s mostly an excuse. What I’m really interested in is why things happen and how to predict what’s next.
I have a habit of pulling on loose threads. Sometimes they morph into a spreadsheet. Sometimes they lead somewhere stranger.
This website is a side effect of that habit.
I work with data, but that’s mostly an excuse. What I’m really interested in is why things happen and how to predict what’s next.
I have a habit of pulling on loose threads. Sometimes they morph into a spreadsheet. Sometimes they lead somewhere stranger.
This website is a side effect of that habit.
I build models, frameworks and investigations that help organisations understand customers, measure impact and make better decisions.
investigations
all →who's about to ask for a mortgage? · training a random forest on 100,000 synthetic customers to spot the quiet signals of intent — and finding that behaviour beats demographics, every single time.openThe £2 Million Marketing Mistake · A fictional retailer spends heavily across paid search, social and email. Which channels actually caused the growth?opennot all customers are equal · a subscription business treated every customer the same. the top 8% were paying for the other 92%.openwho will leave next? · a telco was firefighting churn after it happened. a propensity model moved the fight forward by 90 days.open
notes
all →topics
browse →customer intelligence · Understanding who customers are and how they behave.opensegmentation · Grouping customers in ways that drive decisions.openpersonalisation · Tailoring experiences without losing scale.openmarketing measurement · Quantifying what marketing actually does.openattribution · Crediting the channels that earned the outcome.openexperimentation · Learning by deliberately changing things.openforecasting · Estimating what is likely to happen next.opencausal inference · Separating correlation from cause.openpropensity modelling · Predicting who is likely to act.opencustomer lifetime value · Valuing relationships, not transactions.opensingle customer view · One record per person, finally.opendecision science · How organisations choose under uncertainty.open
a publication about decision-making, customer behaviour, and the stories hidden inside data.