Crunching Data & Insights: How Sweet Karam Coffee Savoured QComm Success

Sweet Karam Coffee (SKC)’s revenue growth since its inception in 2015 followed a familiar pattern seen in digital expansion - a slow climb at first, followed by a rapid surge to the peak, only to dip and plateau. That is, until QComm changed the game in 2023.

“We didn't realize that we had built a solid brand. The overall perception of the brand was much bigger than what the revenue figures were talking about. There was so much customer love,” gushes Veera Raghavan, CMO of Sweet Karam Coffee, as he talks about the near-immediate success that the brand experienced on QComm.

While the South Indian snack maker’s success still relied on the pull from the Southern states, they were determined to establish themselves in the North too.

When the brand first launched on Blinkit, they noticed that their SKUs weren’t available across all dark stores in the Delhi-NCR region. To get to the root of the issue, they turned to GobbleCube’s insights and uncovered an unexpected truth—availability wasn’t the problem. The real challenge was listing. Their SKUs were listed in fewer than 50% of dark stores in Delhi-NCR, limiting their reach and potential sales.

Enabled and supported by the platform, SKC further doubled down on data and insights to fast-track its expansion and growth in the region.

“The platform was also happy to see our focus on analytics and the external investments we were making in it,” says Raghavan.

Today, the brand’s dark store presence on Blinkit has skyrocketed from 50-55% to 85-90%, driving an explosive 2.5x growth in Delhi-NCR alone.

In just months, SKC transformed from a pure-play D2C brand to a QComm power-play - proving that with the right data, strategic execution, and market intelligence, scaling isn’t just a goal. It’s an inevitability.

Raghavan spoke to us at length about the brand’s transition from D2C to QComm and its challenges, matching demand projection with availability, and how data and insights formed the core of their QComm success.

Some snippets from the conversation:

The first three months of our QComm journey (early 2024) were a bit of a struggle. We were trying to understand the dynamics of the ecosystem - how the entire model works, the brands, the categories, the price point, products, all of it.

With a step-by-step approach, we established our presence on the three platforms. Our first breakthrough happened on Instamart in April 2024 when we got the right category manager. Then we went live with Zepto and Blinkit soon after that.

To adapt to QComms, we have had to reset our warehousing and back-end capabilities from D2C to QComm. While D2C is order-based delivery, QComm needs longer term inventory planning.

But it has been an incredible journey for us. Today, QComm and D2C each contribute 40-45% to our revenue. QComm has achieved in 8-9 months what took us 7-8 yrs on our D2C platform.

Of course, a huge portion of this QComm growth can be attributed to our loyal D2C customer-base. They carried over the demand to QComm platforms, giving us a ready-made customer-base on QComms.

Which is why 40% of our QComm revenue is organic. We are already seeing roughly ~5.5 lakh brand searches on all three platforms.

To us, QComm has been the most efficient distribution partner, amplifying whatever we were building and giving us the boost we needed.

Having been a D2C brand for nearly 8 years, the DNA of the organization has always been D2C first, from a pricing, marketing, and distribution standpoint.

We used to have not more than 7-10 days of stock whereas on QComms we had to maintain a larger inventory depending on the norm. So at our end, the entire demand projection piece was getting difficult. In D2C my working capital is only for two days whereas on QComm it could be 14-20 days on an average based on the payment cycle.

So even at a P&L level we had to rejig the way we managed our receivables and payables and by extension our entire production.

But things are much better now, given that we have 8-9 months of experience on the platform and also we get more data from the platforms than we did around 4-5 months back.

Our shelf life of 4-6 months is also a constraint for demand projection. At the time of sale, the product should have a freshness of about 70%. That means I cannot stock up too much at my end.

So I can only do weekly or bi-weekly production planning. This is where our entire demand projection becomes extremely important.

Because if I over produce, then I run the risk of overstocking and rejection at the time of the GRN because the freshness falls below 70%. Of course, our D2C channel allows us to exhaust any stock spill over that may happen.

But if I am measured in production, I run the risk of understocking and going out of stock. So demand projection becomes tricky.

We have been able to scale 20-30% m-o-m on QComms while keeping our marketing costs optimal, thanks to the insights we have been mining using GobbleCube.

Before GobbleCube, availability or dark store penetration was a black box to us.

But with GobbleCube we know our daily sales, darkstore penetration, darkstore availability, stock levels, days of inventory - which helps resolve our demand projections issues to a large extent. We know how much we have supplied, the daily sales data, the DOI and so we can plot realistic demand projections for the next 7 or 14 days and even the next 12 months.

Now, that is also helping us plot our projections against the actual sales that we have achieved, which I would say is almost 90 to 95% accurate.

Often we underestimate our own SKUs because of a lack of data and insights. For instance, the Madras Mixture was selling about 100 or 125 units daily in Chennai which was a good number for us. But with GobbleCube data we realised that this was way below its actual potential. With a proper gap analysis based on the sales data, SOH or DOI levels, we were able to arrive at the true potential of the Madras Mixture in Chennai.

On the marketing front, the ad SOVs that we get have helped us optimize ad spends better by recognizing the headroom available for marketing budgets. By mapping ROI to the sales and ad spends, we have been able to identify the success or under-utilization of our ad spends and actually drive up sales.

Earlier, if sales were down, we knew it was because either my marketing budget or availability had dropped. But with data on our category share, a drop in revenue can then be attributed to a shift in platform strategy or competitor strategy.

For instance, during Valentine’s day, the chocolate category does better which is why my revenue falls on that day. But my category share remaining intact tells me that things are stable at our end.

Now let's say my category share falls but for a couple of my competitors, the share goes up. GobbleCube data clearly tells me that a certain brand has increased their discounting percentage by 10-20% because of which my revenue is falling and my share is going to another brand. The beauty of this is that I get all this without seeing my competitors’ revenue data.

GobbleCube also tells me the revenue percentage split between Zepto Super Savers and non-Super Savers. This helps in very specific decisions regarding discounting terms for Super Savers vs non-Super Savers.

Given that we principally believe in building the brand and not a discounting-led or heavy marketing spends approach to grow our sales, these insights have contributed significantly to our growth.

In the first six months on QComms, there is already a delta sitting for you to immediately enjoy. But that also means that you have to win their (QComm platforms) confidence. To do that, you need to achieve a certain fill rate, have your inventory sorted, etc.

We have been honestly telling platforms that we use GobbleCube. That also signals to them that we as a brand are actually interested in growing on the platform by leveraging analytics (apart from what we already get in the platform). Otherwise, we are manually checking 10 or 15 pin codes and then going to them with availability concerns.

But if platform data shows availability at 85-90%, then we can’t do anything about it.

However, things are far more concrete now with us being able to talk about darkstore penetration, stock levels, DOI, etc. As we are showcasing more data-led decisions, platforms acknowledge it and extend support wherever possible depending on the traction.

For them too, their space is gold. Because we are visibly growing on all the three platforms, they have been more than happy to understand data and support us in these course corrections. And that’s where the actual unlocks have been happening.