The Information Stack, explained
Sriram ChidambaramFounder & Managing PartnerIf the financials aren't the system, what is? The whole Information Stack framework in one place: three data streams, five layers, and the discipline that turns raw data into decisions and capital.
I spend a lot of my time telling founders that their MIS is not their P&L — that the they call "management information" are really just the exhaust of the business, produced a month too late to do anything about. It's a useful thing to hear. But it raises an obvious question, and it's the one I want to answer here. If the financials aren't the system, what is? What are you supposed to build instead? This is that answer. It's the framework I've arrived at over twenty-six years around finance and close to a decade sitting with growth-stage founders, and I call it the Information Stack. It isn't a reporting tool or a dashboard. It's the way information should move through a company — and, built well, it's the difference between a business that can see where it's going and one that's flying blind.
Start with the cockpit
The image I keep coming back to is a cockpit. A company flies on its instruments. Not on the view out of the window, and certainly not on the pilot's gut — on a trusted set of readings, fed by sensors all over the aircraft, that tell you the true state of things in time to act. When your fuel runs low or you drift off course, a good panel shows you the moment it happens. You correct. You don't wait for a report next month. The Information Stack is that instrument panel. And here's the uncomfortable truth it exposes: a great aircraft with instruments you can't trust is not a great aircraft. It's a crash waiting for weather. Most growing companies aren't short of engine power — they have a good product and a real market. What they're short of is a cockpit they can rely on when things get bumpy. Building that cockpit is the whole job. So how do you build it? The Stack runs on two axes. Get these two ideas and you have the whole framework.
Axis one: the three data streams
Strip away the industry and the jargon, and every organisation on earth runs on exactly three streams of data. There is people data — who you employ, how they're organised, what you pay them, how they perform, and why they leave. There is sales data — your pipeline, your customers, your revenue, and how you go to market. And there is operations data — how you actually make and deliver what you sell: your supply chain, your inventory, your product, your fulfilment. That's it. Everything else is a combination or a consequence of these three. Now here's the part that reframes how you see your own business. Your financial statements are not a fourth stream. They are simply the place where the other three meet and turn into money. Sales becomes revenue. Operations becomes cost. People becomes payroll. The P&L is the three streams, translated into a common currency, one period after the fact. Which means a founder who watches only the P&L is managing one-third of their own information — the sales stream, usually, because that's what everyone watches — and seeing it too late to steer. It's like flying a three-engine plane while looking at a single dial. The other two engines are running whether you're watching them or not.
Axis two: the five layers
The three streams are the inputs. But raw data isn't worth anything until it's been through a process — and that process has five stages. These are the five layers, and every stream has to climb all of them. Source — one version of the truth. The foundation. Do your systems agree on what a number even means? When "active customer" or "booking" resolves to three different values in three different tools, you have a broken Source layer, and everything above it is built on sand. This is where most fights that look like reporting problems actually live. Measurement — the right instruments. Once the base is trusted, what do you actually measure, and why? This is where vanity metrics get replaced by the handful of numbers that change a decision, and where an operational movement gets connected to its financial consequence. Meaning — the story behind the numbers. A report tells you what happened. Meaning tells you why, and what it implies for next quarter. It's variance explained by cause, a rolling forecast instead of a stale annual budget, and tradeoffs named out loud rather than hidden. Decision — information built to act on. This is where the board pack lives, and the test is brutal: could someone allocate capital from this page without a meeting to explain it? Most packs fail — too long and too thin at once — and the business learns things too late as a
result. Confidence — the face you show the world. The top layer: your investor updates, your capital narrative, your data room. It's where a clean base finally pays off, because the numbers bridge and hold up when someone looks closely. And it's where a weak base finally gets priced. Read the layers from the bottom up and you see how a company earns capital: trusted data becomes good measurement, becomes honest analysis, becomes sound decisions, becomes confidence. Read them from the top down and you see how an investor's question decomposes into the operational fact that answers it.
Put the axes together
Here's the move that gives the framework its power. You don't have "a Source layer" in the abstract. You have a Source layer for your people data, another for your sales data, another for your operations. Each of the three streams has to climb all five layers. Which means the real diagnostic question about any company is never "do you report?" It's "which stream, at which layer, is weakest?"
Two rules that govern the whole thing
If you remember nothing else about how the Stack behaves, remember these two. Architecture fails from the bottom up. A stack is only as strong as its lowest broken layer. A company with a brilliant Meaning layer built on a broken Source layer is more dangerous than one that's weak everywhere, because it will present with total confidence on numbers it cannot actually defend — right up until diligence. So when you assess yourself, don't average your score. Find your weakest load-bearing layer, because that's the one quietly doing the damage. The disease is imbalance, not absence. Don't ask whether you have reports. Ask whether information flows, trusted, from the shop floor all the way to the cap table without changing meaning along the way. That's the real bar, and almost no growing company clears it without deliberately building for it.
What this looks like in a real company
Let me make this concrete, with a company that's a composite of ones I've sat with. Call it a Series B business — good product, real growth, a founder who came up through sales. Their sales stream is genuinely strong. Clean pipeline, an ARR dashboard the whole team
watches, conversion tracked stage by stage. On our model, that stream is at Level 4 — it's what won them their Series A and most of their B. But look one layer down, and across. Their Source layer is cracked: the CRM, the finance system, and the founder's own board slide each define an "active customer" slightly differently, so "revenue" quietly means three things. Their people stream is barely instrumented at all — headcount in a spreadsheet, attrition discovered the day someone resigns. And nobody's watching cost-to-serve, so margin is a guess. Average all of that out and you'd call them a healthy Level 3 company. That average is exactly the trap. Because architecture fails from the bottom up, their real level is set by that cracked Source layer and that ground-floor people stream — they are a Level 1-to-2 company presenting like a Level 4. And it held together right until it didn't: a diligence process that dragged for weeks because the revenue numbers wouldn't bridge, and two senior departures in a single quarter that the data had been signalling for months, if anyone had been reading it. Nothing was wrong with the business. Everything was wrong with the cockpit.
One stack, many shapes
The five layers never change. What changes, by business type, is which stream binds first and how you instrument it. For a SaaS company, the hardest work sits at the Source and Measurement layers of the sales stream — nailing the definitions of bookings, billings, and recognised revenue, and then retention and on top. For a D2C brand, the centre of gravity shifts to operations: fulfilment, inventory, and worked out cohort by cohort. For a manufacturer, it's the inventory chain and the translation of throughput into margin. And for a hospital or a diagnostics lab — one of the most information-hungry businesses there is — it's a revenue chain split across payers who each behave completely differently, sitting on top of capacity economics like bed occupancy and machine . Same cockpit, different instruments. The stack is universal; the gauges are specific. Knowing which stream will hurt you first, given what kind of company you are, is most of the art — and it's why the same diagnosis produces very different first moves.
Who owns the truth
There's one structural point I feel more strongly about than almost anything else, because it's where good intentions go wrong. The Information Stack — the single, reconciled version of the truth — has to be owned by someone neutral. In practice that means it reports to finance, or in a smaller company, to the founder directly. Never to the function whose own
numbers it measures. The logic is simple and unforgiving. If the sales team owns the revenue data, revenue will drift optimistic — not through dishonesty, but through the natural pull of incentive. If operations owns the delivery data, delays get quietly softened. If a business unit owns the numbers that set its own budget, those numbers will bend, as surely as water runs downhill. The keeper of the truth cannot be someone with a stake in what the truth turns out to be. A cockpit whose gauges are wired by the very people they're meant to check is a cockpit that will, sooner or later, fly you into the ground. Neutral custody is what keeps the instruments honest.
One source, many altitudes
The last idea ties it all together, and it's the one that saves companies from their most common self-inflicted wound. Your Stack doesn't serve one audience. It serves several — your management team, your board, your investors, your bankers, your auditors — and each of them looks at the business from a different height. Your board is inside the aircraft and wants the operating detail and the honest problems. Your investors want the durable trajectory, the risks, and the ask. Your banker wants the cash and the cover. A mature Stack does not build a different version of the truth for each of them. It builds one trusted foundation and renders it at different altitudes. Same source, different altitude. The moment you start keeping different underlying numbers for different rooms, you've stopped building an information system and started building a liability — one that will surface at the worst possible moment, usually in a data room, and usually be read not as an error but as a verdict on you.
Where to start
If all of this feels like a lot to build at once, here's the mercy in it: you don't have to. Because architecture fails from the bottom up, you don't fix everything — you find your single weakest load-bearing layer, in your weakest stream, and you start exactly there. Usually that's less glamorous than founders expect. It's rarely a better forecast. It's far more often locking a handful of definitions so that "revenue" and "active customer" finally mean one thing, and warming a data room that's gone cold. Fix the floor first; the rest of the climb gets easier once the base holds. Working out where your floor actually is — which stream, which layer — is the one thing worth doing before anything else. It's exactly what a proper diagnostic is for.
The whole thing, in a sentence
So here is the framework, complete. Three streams — people, sales, operations — climbing five layers, from a trusted source of truth up to the confidence that moves capital, held honest by a neutral owner, and shown to every audience from one foundation at the altitude that suits them. That's the cockpit. Build it, and you stop reacting to last quarter and start flying on live instruments — faster, further, and straight through weather that grounds everyone else. Your financial statements will always tell you what already happened. The Information Stack is how you get to shape what happens next. Sriram Chidambaram is the Founder and Managing Partner of SRF Capital Studio, where the Growth Stage CFO Office helps founders and MSMEs build the Information Stack — finance as an information function, not a reporting one.
How relevant and useful is this article for you?
The next one
Get what we publish next, by email.
Working notes on raising, borrowing, protecting, growing and structuring capital in India. One email a week at most, and you can leave any time.
We use your address only to send this. See our privacy policy.
We store your address to send you these emails and nothing else. See our privacy policy.
Related reading
