Consulting · Ethnic and bridal wear · Mumbai
Seventy-five roles, none of them written down.
The short version
When we went looking for KALKI’s data in 2018, there was almost none. A fast-growing bridal and ethnic wear retailer was making its decisions on experience and instinct, which had worked well enough to get it that far. Over eight months we documented75roles across every function, gave each one indicators and a tracker, and built dashboards at four cadences for every store, department and process. Five years later they brought us back for two months to rebuild the store checklists — because the documents that tell you how to run a store are not the documents that tell you how to open one.
There was almost nothing to analyse
Saurabh, Shishir and Nishit Gupta brought us in to structure the internal systems behind the brand — productivity, operational efficiency, and data tracking across the organisation. A confident brief from a management team that could see what was coming.
We began the way we usually do, by asking for the data. There was very little of it. What existed was partial, and most decisions were being made on historic experience or on an individual’s read of a situation.
That is not a criticism of anybody. It is what almost every fast-growing business looks like before someone sits down and decides what is worth counting. But it does set the order of work: you cannot improve performance you cannot see, and you cannot build a performance system on top of nothing.
A business running on judgement is running on the judgement of the specific people who have it. That scales exactly as far as their attention does, and no further — which is fine at four stores and impossible at fifteen.
Working out what was worth counting
We ran detailed sessions with each member of the management team separately, on what they wanted this exercise to have fixed by the end of it. Asked separately, senior people give different answers, and the differences are informative.
Then we sat with every unique role in the organisation — not every person, every distinct role — and went through what the job actually involved, where it got stuck, and what the person doing it thought would make it work better. Seventy-five of them.
Finally we went back to management with a two-to-three year picture of where they intended to be, and built the project backwards from that. Indicators chosen against where a business is going are useful. Indicators chosen against where it currently is mostly measure the past.
What we built
| 01 | Mission, vision and values, defined with the management team — the reference point for every subsequent decision about what to measure and reward. |
| 02 | Job descriptions for all 75 roles, each with its own indicators and its own data tracker, explained individually to the people holding them. |
| 03 | Checklists, processes and workflows for every function, with training run for every individual rather than a document circulated by email. |
| 04 | Dashboards at daily, weekly, fortnightly and monthly cadence — for each store, each department and each process, built around business objectives rather than around what happened to be easy to collect. |
| 05 | HR policies reworked and an employee handbook written and communicated across the organisation. |
| 06 | An incentive and reward programme for the sales team. |
| 07 | Daily huddles across every store and department, reviewing the previous day’s wins and lost opportunities. |
| 08 | Recommended changes to the in-store customer journey at the points where it was working against the sale. |
The huddle that trained the room
One design decision is worth pulling out, because it is the part most likely to be copied badly.
The daily huddles were not led by the store manager. They were led by a different person every day — whoever’s turn it was ran the review, presented the numbers and chaired the discussion of what had gone wrong yesterday.
A daily review meeting run by the same person every day produces a daily review meeting. Rotate who runs it and you get the review plus a year of quiet practice in reading numbers, speaking to a room, and being accountable for something in front of colleagues. The second one is worth more than the first.
It also does something a training budget cannot: it identifies who can do this. Within a few months of rotation, a management team knows which of its people can hold a room and think on their feet, and it knows it from evidence rather than from impression.
Running a store and opening one are different problems
In April 2023 KALKI came back with a specific and time-boxed question. They were about to open stores across India at a pace they had not attempted before, and they wanted every store checklist reviewed before they did.
The checklists from 2018–19 were working. But they had been written for the stores that existed, by people who knew those stores — operational documents for running a shop that was already running. A rollout needs something different in three ways:
| They have to work for a store nobody has operated yet | An operational checklist assumes context that a new team does not have. |
| They have to flex by store size | A process built around a large flagship breaks in a smaller format, and a process built around a small store leaves a flagship under-managed. |
| They have to cover opening, not just operating | Everything between an empty unit and a trading store had never needed to be written down, because it had only ever been done by the people who did it. |
Two months, every store-facing checklist reviewed and reworked against those three requirements. It is unglamorous work and it is the difference between a rollout that scales and a rollout where every new store is a bespoke rescue operation.
What we can and cannot tell you
| What we can evidence | Figure | Basis |
|---|---|---|
| Roles documented | 75 | Every unique role across every function, each with a job description, indicators, checklists and a tracker. Our own project records. |
| Functions covered | all | The first engagement covered the whole organisation rather than a department. |
| Dashboard cadences | 4 | Daily, weekly, fortnightly and monthly, built for each store, department and process. |
| Duration | 8 months | September 2018 to April 2019. The second engagement ran April to May 2023. |
| Still in use | yes | The systems built in the first engagement are still the ones the business runs on, and were the basis the store playbook was rebuilt from in 2023 rather than being replaced. |
How to read this
We had no access to their financial performance. Consulting engagements of this kind do not usually come with the client’s internal numbers, and this one did not. We can tell you what we found, what we built, how much of it there was and whether it is still in use. We cannot show you a return, because we were never in a position to measure one.
The company’s growth is not our claim. KALKI has grown a great deal since 2018, and the reasons are overwhelmingly to do with the brand, the product and the people running it. What is set out below is narrower and is the only thing we would put our name to.
The most persuasive evidence here is that they came back. A client who returns after four years, for the specific purpose of rebuilding documents we originally wrote, is making a judgement about that work with their own money. We would rather rest on that than on a number we cannot source.
What happened next
The rebuild was done in April and May 2023 for a rollout that was about to begin. It began.
By December 2023 KALKI was operating five stores across Mumbai, Delhi, Ahmedabad, Bangalore and Surat, with more than 35,000 square feet of retail space and four further markets identified. By 2025 it was reporting eight flagship stores, a 6,000 square foot flagship had opened in Mumbai, and a Chennai store followed in January 2026. In April 2025 the business raised₹225 crorefrom Lighthouse Funds to fund exactly that expansion.
We are not claiming the rollout, the funding or the growth. What we can say is narrow and checkable: the store playbook that rollout ran on was rebuilt for the purpose two months before it started, on top of systems built five years earlier that had never been replaced.
Constraints that shaped the work
Starting with no data means the first months produce nothing anyone can see. Trackers get built, indicators get argued about, and the dashboards stay empty until enough has been collected to be worth looking at. That is a long time to ask a management team to keep faith in a project, and the ones who do are the reason it works.
The second engagement was two months, which is enough to rebuild documents and not enough to see them used. We reviewed and reworked every store checklist against a rollout that had not happened yet. Whether they held up in a store nobody had operated before is something the client learned after we left, and we would want a review built into the next one of these rather than a handover.
In their words
“They really helped us structure all our reports and also give a defined checklist and tasks for every team. Because of that, we can now quantify the work done in every department — it has really helped us in improving our efficiency and productivity. They took tremendous effort in understanding how our business works and customise reports and requirements as per our needs.”
Ritika Gupta — Director, KALKI Fashion
If you are about to scale a store network
Check what your store documents actually are. Most retailers have operational checklists written by people who know the store — useful for running what exists, close to useless for opening what does not. The gap only shows up at the third or fourth opening, by which point you are running the rollout on the personal attention of whoever wrote them.
Rewriting them before the rollout costs a couple of months. Discovering the problem during it costs considerably more.
Read another
Every figure sourced from the client’s own systems, with the basis stated and the exclusions named.Bawarchi Atlanta
Smile Loft Dental
Sanpeggio’s Pizza
Measurement note. Role counts, scope and duration are from our own project records. Whether the systems remain in use is confirmed by the client. Store counts, retail footprint, revenue and funding are drawn from published trade press and third-party company records rather than from the client, and are included as context on what the business did after our involvement, not as outcomes we are claiming.
