Case Study 2026

Small businesses are flying blind on social media.

Every week, owners watch their engagement rise and fall with no idea why. Pulse is a diagnostic dashboard that names the cause, ranks the fix, and turns daily anxiety into a daily decision.

Diagnostic UX Data Storytelling Small Business Mobile-first
Detected
Drop in Reels output
0 Reels this week — your top format.
Needs Attention
Engagement down 34%
Likely cause: content mix shift. Recovery in 7–10 days.
Reach
24.8K▼28%
Engagement
2.4%▼34%
Visits
1.2K▲8%
Recommended
Post 3 Reels this week
+38% est. reach · 7 days to recover
01 / Context

The feed is now the storefront.

A missed week on Instagram doesn't just feel bad — it costs orders. For independent founders running everything themselves, the social channel is where customers discover, judge, and decide. So when the numbers wobble, the anxiety is real.

Most owners we spoke with run their feed between everything else — packing orders, replying to DMs, opening up the shop. They post when there's time, hope it lands, and glance at the metrics later.

When something dips — reach falls, comments dry up — they notice instantly. The trouble starts at the next step. There's no obvious way to tell whether the cause is content, timing, algorithm, or audience. So they guess. And guessing is exhausting.

“ My engagement dropped 60% last month and I genuinely have no idea if it was my content, the algorithm, or just bad luck. I'm flying blind.
02 / Problem

They can see the drop. They can't see the cause.

Native analytics tools tell business owners what happened, but never why. The result: anxious decision-making based on hunches, not data.

?

Is it the content? Did the last few posts not land — or is the format the issue?

?

Is it the timing? Posting at the wrong hour, the wrong day, or too infrequently?

?

Is it the algorithm? Did Meta change something behind the scenes again?

?

Is it the audience? Are followers losing interest — or has the demographic shifted?

03 / Goal

Detect. Diagnose. Decide.

A dashboard that does the analytical work for business owners — surfacing problems, explaining them in plain language, and recommending the next move.

01

Detect drops fast

Surface engagement issues the moment they happen, before momentum is lost. Smart anomaly detection that watches the numbers so owners don't have to.

02

Diagnose the cause

Compare content, timing, format, and audience signals to pinpoint what changed. Translate the analysis into clear, jargon-free explanations.

03

Recommend action

Suggest specific, prioritized fixes — not a list of metrics to interpret. Tell users what to post, when, and why it will help.

04 / Research

What we learned from real users.

A mix of generative and evaluative methods — going beyond traditional interviews to understand actual behavior, not just stated preference.

01.

Jobs-to-be-Done Interviews

12 small business owners interviewed about their relationship with social media — focused on motivations and outcomes, not demographics.

Generative
02.

Diary Studies

5 participants logged their social media check-ins for 2 weeks. Surfaced the emotional patterns and decision moments behind the data.

Behavioral
03.

Competitive Heuristic Audit

Evaluated 8 competitors — from Meta Business Suite to Sprout Social — against Nielsen heuristics and JTBD coverage.

Evaluative
04.

Co-Design Workshops

Two remote sessions with 6 owners using FigJam — they sketched their ideal "what's wrong" report. Patterns emerged fast.

Participatory
05.

Usability Testing

Moderated sessions with prototypes at three fidelities. Tracked time-to-insight, confidence ratings, and confusion points.

Validation

Key findings

11/12

Check their analytics daily but admit they "don't know what to look for"

4×

The average number of tools owners switch between to piece together a picture

73%

Have changed their strategy based on hunches in the last 6 months

$38

Average willingness-to-pay per month for a tool that diagnoses problems

05 / Users

Two owners. Two very different needs.

The research surfaced two distinct user types — both small business owners, but with different relationships to data, time, and risk. Designing for both meant balancing simplicity with depth.

Sarika Sharma portrait
Sarika Sharma
The Hands-On Founder
Age32
BusinessFashion DTC
Followers3.2K
Context

Runs her sustainable fashion boutique solo. Posts 4–5 times a week and checks Instagram between customer DMs and inventory runs. Confident with content but uncertain with numbers.

Primary need

A quick read on what's going wrong — without having to learn marketing terminology. Tell her what to fix and she'll fix it.

"I don't need more dashboards. I need someone to tell me what to do next."
mobile-first low-data-literacy time-poor action-oriented
Mihir Patel portrait
Mihir Patel
The Growth-Curious Operator
Age41
BusinessCafé (2 locations)
Followers18.5K
Context

Manages marketing for both café locations with a part-time helper. Comfortable with metrics, runs occasional paid campaigns, but still feels he's missing the "why" behind organic dips.

Primary need

Diagnostic depth — he wants to understand the reasoning, drill into segments, and test hypotheses. A tool that respects his data fluency without burying him in spreadsheets.

"Show me the diagnosis, but let me see the working. I want to know why."
desktop+mobile data-fluent strategy-focused curious
06 / Wireframes

Mapping the diagnosis flow.

The core experience had to compress hours of analysis into a single scroll. Early wireframes focused on hierarchy — what the user sees first, second, third — before any visual treatment.

07 / Iterations

Three rounds. Three turning points.

Each design iteration was tested with 4–6 participants from the target personas. The pattern was consistent: less is more, language matters, and mobile is non-negotiable.

PULSE / OVERVIEW · LAST 30 DAYS · ALL ACCOUNTS IMPRESSIONS 142,318 ▲ 4.2% REACH 88,402 ▼ 12.8% ENGAGEMENT 3.18% ▼ 0.4pp FOLLOWERS 12,481 ▲ 1.1% CTR1.92% SAVES421 SHARES187 COMMENTS96 PROFILE VISITS2,103 LINK CLICKS312 VIDEO VIEWS8,442 AVG WATCH4.1s FILTERS: RANGE ▾ PLATFORM ▾ CONTENT TYPE ▾ AUDIENCE ▾ COMPARE ▾ "where do I even start?"
Prototype · v1 · Figma
The "everything dashboard."

12 metric tiles, 3 filter rows, no hierarchy. Designed to impress in a demo — froze users in a real session.

12 metrics no verdict desktop-only
v1

The Insight

"There are too many numbers."

The first prototype tried to be comprehensive — 12+ metrics visible at once. Sarika opened it, paused, and said "I don't know where to start." The mistake was treating the dashboard like a report.

The Change

Hide depth behind a click.

Reduced the top-level view to three metrics and a single status signal. Detail moved into progressive disclosure, available but not assumed. Mihir could still drill in; Sarika didn't have to.

The Result

Faster to first insight.
−40% time to identify a problem
PULSE / THIS WEEK NEEDS ATTENTION Your posts are reaching fewer people than usual. PEOPLE WHO SAW YOU 88k down 13% vs last week PEOPLE WHO REACTED 2.8k down 8% NEW FOLLOWERS +138 up 1% See what's causing the drop → "oh — I get it."
Prototype · v2 · with copy pass
Plain language, one verdict up top.

Every label rewritten conversationally. The dashboard now opens with a sentence, not a stat block. Detail still one tap away.

3 metrics verdict-first copy-led
v2

The Insight

"What does 'impressions' mean again?"

Even our cleanest layouts failed when the copy was technical. Testers visibly tuned out on terms like "reach vs. impressions" or "engagement rate by impressions." The numbers didn't matter if the labels were opaque.

The Change

Rewrite everything in plain English.

Worked with a content strategist to translate every label, tooltip, and recommendation into conversational language. Replaced jargon with action: "Your posts are reaching fewer people than usual."

The Result

Higher confidence in next steps.
+40% comprehension scores
MOODBOARD · THUMB-FIRST one-handed in-line swipeable cards thumb zone trend band REF · IG STORIES · LINEAR REF · ARC SEARCH REF · CASH APP CARDS NEEDS ATTENTION Reach is down this week. FIX #1 · HIGH IMPACT Post at 7pm IST FIX #2 More reels, fewer carousels SHIPPED SUS 89 mobile THUMB ZONE bottom 60% CARDS swipeable
Moodboard + Prototype · v3
Mobile-first, thumb-native.

References pulled from IG Stories, Linear, Arc Search and Cash App. Diagnosis becomes a vertical scroll; fixes become swipeable cards.

thumb zone cards, not tables one-handed
v3

The Insight

"I check this from my phone in line at the bank."

Diary studies revealed the real use case: opportunistic mobile check-ins, often one-handed, often interrupted. Our desktop-first design assumed focused time the users didn't have.

The Change

Mobile-first restructure.

Rebuilt the diagnosis flow as a vertical scroll, optimised for thumb reach. Charts simplified into trend bands. Recommendations turned into swipeable cards. Desktop became the secondary experience.

The Result

Production-ready usability.
89 SUS score (mobile)
08 / Final Designs

What shipped — and why.

The final product is built around four principles. They're not novel, but they're applied with intent — every screen, every label, every interaction.

The thesis

A dashboard that answers, not one that asks.

Open the live prototype
  1. 01

    Diagnosis-first

    Lead with what's happening and why — not with a wall of metrics to interpret.

  2. 02

    Plain language

    Replace jargon with sentences a small-business owner reads once and understands.

  3. 03

    One action per screen

    Every view points to a single next step, so decisions don't pile up.

  4. 04

    Mobile-native

    Designed thumb-first — the phone is where most checks actually happen.