The Sourdough Principle
Or why research should compound, not expire
When I was growing up, one of the small kitchen miracles that blew my mind was curd. Yes, curd - that staple of Indian households. My mom would make it afresh daily, but she would never use all of it - a small portion was always kept aside. As an inquisitive kid, I asked her why she did this. She explained that this saved curd would be added to milk to create… more curd. And there it was the next morning - a vessel full of fresh, yummy, creamy, curd. To a child, this felt like alchemy.
Maybe it was then, in the mind of a curious 8-year old, that the idea that the old could generate the new was seeded.
Our obsession with the new
Not everything valuable needs to be created from scratch - it is likely that what you already have has plenty to offer. In the modern corporate world, I feel research especially suffers from this obsession with the new.
A new brief comes in, a new question is asked, a new study is commissioned, a new sample is recruited, a new deck (or ten!) is made, a few decisions are taken. Then the work is filed away, perhaps never to be used again. A quarter later, a few urgent questions arrive. The cycle repeats.
This is understandable. For a long time, it was also practical. Even if the old intelligence existed somewhere, using it well was hard. It might be scattered across decks, transcripts, dashboards, old qual, survey reports, social listening, brand trackers, strategy documents and market studies. The information was technically available, but it was not decision-ready. Since it was purpose-made to answer last quarter’s question, surely it would be too much to ask it to answer this one’s.
New also has its own glamour. New feels current, safer, like action. But new also has a problem: it often starts from zero.
Our brains rarely work that way. When a strategist, researcher or founder encounters a new question, they do not begin with a blank slate. They instinctively reach into memory. They pull from past projects, old conversations, consumer contradictions, category patterns, market shifts, things they have seen before but cannot fully cite. They test a scenario using what they already know before they go looking for more.
Human judgement is always compounding. Organisational intelligence often does not.
For years, the logic made sense in theory but was difficult to build in practice. How do you take years of accumulated research and make it usable again? How do you ask a new question of an old interview? How do you connect a 2021 segmentation, a 2023 qual study, a dashboard, a strategy deck, and a new market signal without turning the exercise into a smorgasbord of summaries?
And then came AI
When generative AI entered the room in late 2022, I saw some potential. Not because ChatGPT could suddenly “do strategy.” It cannot, by itself.
But because it could change the economics of working with accumulated knowledge. It could retrieve, connect, compare, compress, interrogate and simulate at a scale that was previously unrealistic. More importantly, when designed with the right strategic guardrails, it could make old intelligence queryable again.
Digital twins now become interesting. A good digital twin is not a fictional persona generated from the internet. It is not “Urban Priya, 29, loves convenience and authenticity”. A useful digital twin is built from real source material: transcripts, verbatims, behaviours, attitudes, choices, contradictions and context. It allows teams to ask new questions in the vicinity of what the original data can support.
Questions such as: How might this respondent react to a new product idea? Would this segment see a premium extension as credible? What barriers would this consumer notice first? What language would feel natural?
The twin is not the answer. It is an instrument.The strategist still decides what to ask, what to trust, what to challenge, and what to take forward.
That is what we call the Sourdough Principle.
It’s what my mom told me about curd - which extends to kombucha, kimchi, and of course, everyone’s favourite pandemic-era bread.
Every organisation already has starter material: past research, consumer conversations, strategy work, category understanding, failed ideas, old wins, abandoned territories, customer journeys, behaviour patterns, market observations. On their own, these assets may look like leftovers. But when they are fed with new questions, fresh context, and strategic judgement, they begin to compound.
The Sourdough Principle has four parts:
Over the last year of building Qlue, I and my co-founder Ravi have refined this philosophy and put it to test in the real world, with some pretty big brands. Along the way, we’ve come to realise that the Principle not just has merit, but has nuance. We’ve realised it has four parts:
- The archive is not a filing cabinet. It is a living asset.
- Fresh input still matters.
- Connections are often more valuable than summaries.
- Human judgement remains the chef.
The Sourdough Principle in action
In one project with a global food brand, this principle was used to explore the future of a beloved dairy brand. The easy answer would have been to commission a long sequence of fresh studies. Instead, the team began by activating years of existing intelligence.
That archive became the starter. Fresh qualitative work then fed the starter. Digital twins helped scale consumer judgement, generating, evaluating and iterating on ideas at speed. Human strategists shaped the questions, interpreted the patterns, merged or discarded territories, and turned the work into an innovation architecture the business could act on. You can read more about this, in our case study here.
The point was not that old research had all the answers. The point was that old research had more to give once it was put back into motion.
That is the shift I care about.
Don’t throw away your starter!
For years, companies have invested heavily in understanding consumers, markets, categories and culture. They have paid for the interviews, run the workshops, commissioned the trackers, built the decks, seen the patterns.
But much of that intelligence is underused because it lives in formats and systems that were designed for delivery, not continuation. The report was the end of the project - while It should have been the beginning of the next one.
The Sourdough Principle asks companies to look again at what they already know - as a serious strategic asset, making the intelligence base richer - just like my mom’s curd - every time it is used.