One designer. Three user types. One broken pipeline. Here's what changed.
Before this product existed, the agency was managing over 1 Lakh in monthly client ad spend — with no system, no pipeline, and no way to know if a lead ever reached the person supposed to call it.
The agency was running paid campaigns for 10+ clients. Leads came in daily. They disappeared into a black hole.
The "system" in practice: a different database with no owner, no timestamp and no status · manual WhatsApp copy-paste by hand · personal phones with no context or history · silent failure, where no one knew when a lead went cold.
In B2C sales, a lead not contacted within 5 minutes is 10× less likely to convert. This agency was averaging two days.
Admin wants monitoring. Client wants transparency. Employee wants speed. All three need different surfaces from the same data. A single shared dashboard solves nothing — it just spreads the confusion more evenly.
Before speaking to a single user, I asked for two days to watch how the current system actually worked. Shadowing revealed what no interview could have surfaced: the system wasn't failing occasionally — it was structurally incapable of succeeding.
Then came 10 interviews across 3 user types in 8 days.
| User type | Interviews | What they said | What they really meant |
|---|---|---|---|
| Agency Admin | 2 | "We need better reporting" | "We have no idea what's working" |
| Agency Staff | 4 | "We spend so much time on admin" | "We are the system — and it's breaking us" |
| Client Reps | 2 | "We need more leads" | "We don't trust the leads we're already getting" |
| Client Employees | 2 | "We try to call as fast as we can" | "We get leads with no context and no urgency signal" |
By the time I get the lead, they've already talked to someone else.
Client employees — the people whose job was literally to call leads and convert them — had no dedicated tool, no context, and no way to signal when they'd called. They were the most important user for the business outcome. They were completely invisible to the system.
This single quote crystallised the entire design brief. Speed wasn't a nice-to-have. It was the product's only real job.
The old system had no structure. Mapping the IA forced every stakeholder to agree on what lived where before a single screen was drawn.
I also pushed back when Dev said it was difficult to secure form data submitted on a single platform — I said it was needed for the workflow.
Six screens. One coherent system. Dark navy sidebar, indigo accent, role-aware surfaces. No spreadsheets. No WhatsApp.
Arjun was flying blind — running 10+ clients with no way to see what was working. The easy build is to put everything on one screen and let him filter. I didn't. I picked the three questions he actually asks every morning and made the screen answer only those.
What that looks like: four number cards with up/down versus last month, a lead-source donut, and a recent-leads list with colour-coded status.
No tables on the dashboard. A manager should see the answer in a chart, not read a spreadsheet and do the math. And a calm blue palette, because this is a screen someone stares at all day.
In the old setup everyone had full access and no one knew who was responsible for what — the same accountability gap that let leads go cold. So on this screen the one decision that mattered was making each person's access an explicit choice at the moment they're added, not a blanket "admin" or "viewer" handed to everybody.
The form walks through it in short steps and ends on exactly what that person can see and do; the team list then shows everyone's role and status at a glance, so responsibility is never a mystery.
Sales reps were abandoning the "add client" form halfway through — the culprit was the email (SMTP) setup, a technical step that looked required and stopped them cold.
The decision: make SMTP optional and tuck it away as "advanced." It's still there for the power users who need it, but it no longer blocks the everyday job of getting a client into the system. That single change brought the drop-offs down.
One smaller call in the same spirit: where a section could sit empty — like products not yet linked — I replaced the blank field with a friendly prompt to add one, guiding the rep instead of leaving them stuck.
The agency's whole problem was information vanishing — leads with no owner, no trail, no accountability. I carried one rule into the catalog to make sure the pattern didn't repeat: every detail you enter when adding a product shows up as a column in the list afterwards. Nothing you fill in ever quietly disappears.
That's also why the product fields go beyond a typical CRM — a reference code (SKU) for client calls, a clear owner, and a public/private setting — because for this agency, who owns it and who can see it were the questions that actually got asked.
The whole platform is built from a handful of the same building blocks, reused everywhere. Twelve of these small patterns across all six screens — every one chosen to make the product quicker to scan and easier to use.
I checked the work two ways: usability testing on the prototype before build, and real usage data after it went live.
| What changed | Before | After |
|---|---|---|
| How fast leads got a response | 24–48 hours | Under 2 hours |
| Leads that actually got contacted | ~30% (estimated) | +42% better |
| Time spent on daily admin | 2–3 hrs/day by hand | Down 65% |
| Leads quietly going cold | Common, and no one noticed | Flagged in real time |
| Task completion in testing | No baseline before | 88% across all 3 user types |
We don't just run your campaigns. You can see every lead, in real time, from the moment it comes in.
Handed off to the dev team in 20 days. Already planned next: a drag-and-drop lead pipeline, client detail pages, global search, in-app notifications, a role-based dashboard, and an activity log.
Six screens, one coherent system.
You've just read how I think — the research, the trade-offs, the decisions I defended and the ones I'd make differently. I'm open to Lead and Senior Product Designer roles, remote, hybrid or in-office. Tell me what you're building and I'll tell you where I'd start.