Product · Profitability Infrastructure

Product · Profitability Infrastructure

Challenges in tracking D2C Profitability and how we solved it at BooleanMaths?

Challenges in tracking D2C Profitability and how we solved it at BooleanMaths?

Most brands were tracking performance either on Meta using attributed ROAS, or on Shopify using total store revenue. At the end of every month, the same brand would run a proper P&L using their bank accounts and invoices to find that their margins were barely 5–6%.

Most brands were tracking performance either on Meta using attributed ROAS, or on Shopify using total store revenue. At the end of every month, the same brand would run a proper P&L using their bank accounts and invoices to find that their margins were barely 5–6%.

Average ROAS seen on Meta

5.7x

Actual Margin measured by the Finance Team

7-8%

Time taken to reconcile the P&L

2 Days

Reason 1

D2C brands are stuck in the same cycle

The pattern was consistent across brands. The ad platform ROAS would show 5.7x. The month would end. The CFO would come back and tell them their margin was barely 7–8%. And there was no way for them to debug where the revenue they had earned was actually leaking away.

They weren't making bad decisions , they were making decisions with incomplete information. Their margins was bleeding to a thousand cuts - shipping, RTOs, repackaging, ad spend across multiple platforms, taxes. But none of that was visible on the dashboard that was driving their daily decisions about which ads to run and which SKUs to push.

"Every morning Meta said the ROAS was 5.7x.
At the end of the month the CFO said margins were 7–8%.
And there was no way to debug where the revenue had actually gone."

"Every morning Meta said the ROAS was 5.7x.
At the end of the month the CFO said margins were 7–8%.
And there was no way to debug where the revenue had actually gone."

"Every morning Meta said the ROAS was 5.7x.
At the end of the month the CFO said margins were 7–8%.
And there was no way to debug where the revenue had actually gone."

Reason 2

Order Placed is the wrong signal

We decided to extend the visibility of the BooleanMaths platform from the web journey all the way through to actual order delivery. The right conversion signal to optimise and measure was not an order placed. It was a delivered shipment. Only at that point has revenue actually been earned. Everything before that is a projection.

We approached this in two phases.

  • The first phase was integrating with the major shipping partners Shiprocket, Delhivery, ClickPost, Quickship, WareIQ, and a few others to pull live fulfilment data directly into the platform.

  • The second phase was building an intelligent platform that could account for all the complexities of shipping, return shipping, and reusing those returned orders.

Reason 3

The complexities we had to account for

Building an accurate P&L isn't just a data integration problem. The operational reality of e-commerce logistics is genuinely complex, and a model that ignored the specifics would be wrong in ways that mattered. We had to ensure that a lot of the customisations that brands needed could actually be handled and we had to provide flexibility for individual packaging, handling & tax rates per SKU. Because the unit economics of a fragile, heavy item are not the same as a lightweight accessory, and the model had to know the difference.

Some of the factors we solved for

Incomplete shipping data

Not all order data was available via shipping integrations. There would always be some failures, or cases where the brand was using an obscure shipping partner.

Varying RTO rates by segment

COD orders had a higher RTO rate of 20–30%, while prepaid orders sat at 4–5%. We had to build an RTO projection model that was accurate to the audience mix.

Not all RTOs are scrapped

For some brands, the reuse rate on returned inventory was as high as 70–80%. For others, as low as 10–20%. The economics of an RTO depend entirely on what happens next.

In-transit projection logic

Shipping takes 2–3 days to mature. Live P&L analysis was simply not possible without projection logic that estimated how many in-transit orders would actually be delivered.

Live COGS from Shopify

COGS data had to be fetched live. For merchants who didn't maintain COGS on Shopify, we added a facility to upload that data directly onto the platform.

Per-product tax brackets

Not all products carry the same tax rate. A luxury item might sit at 18%, another product at 5%. Individual tax brackets had to be applied per product, not blended across the catalog.

Getting to within 1%

Only after building in the logic for all of these scenarios were we able to get to within 1% of the margin that a brand's finance team was reporting at the end of the month.

That 99% accuracy is what enables brands on BooleanMaths to generate CM1, CM2 & margin for each Ad and SKU, in real time using profitability metrics.

The goal was to give brands the same quality of financial insight throughout the month that they previously only had in hindsight. The goal was to match our projections with the truth that will be revealed after the bank statements had been reconciled and the damage was already done.

That's the visibility gap we set out to close. And that's what the profitability module inside BooleanMaths was built to solve.

99% Only after building in the logic for all of these scenarios were we able to get to within 0.5% of what a brand's finance team was reporting at the end of the month.

That 99% accuracy is what enables brands on BooleanMaths to make decisions at an SKU level, in real time using profitability metrics instead of straight revenue metrics.

The goal was never a better dashboard. It was to give brands the same quality of financial insight during the month that they previously only had in hindsight after the bank statements had been reconciled and the damage was already done. That's the visibility gap we set out to close. And that's what the profitability module inside BooleanMaths was built to solve.

Get Started with India's best Attribution Platform today

Install on Shopify

AI powered using

Background