Product thinking / Case study

Sekel Tech Platform - B2B2C SaaS - MarTech

Company: Sekel Tech | Role: Product Manager | Year: 2022–2024

Full product ownership of a hyperlocal marketing and location intelligence SaaS platform. Led a 30+ member team, drove ~30% revenue growth, and established the product organisation’s foundational processes.


Context

Sekel Tech is a B2B SaaS company offering hyperlocal digital marketing and location intelligence to enterprise clients in retail and automotive sectors. When I joined, the product had strong core features but struggled with:

  • Long delivery cycles driven by client-customisation requests
  • No clear roadmap - reactive to whoever shouted loudest
  • Poor sprint predictability - >40% of sprint commitments missed
  • High churn risk on 3 key accounts due to feature gaps

Discovery & First 60 Days

What I did in the first 60 days:

  1. Client audit - interviewed all 10 active clients; categorised issues as: critical bugs, feature gaps, expectation misalignment, UX problems
  2. Team audit - ran retrospectives with engineering, design, QA; uncovered root causes of delivery delays
  3. Revenue analysis - mapped features to revenue: which features were in high-value contracts vs. which were maintenance overhead
  4. Competitive landscape - benchmarked against 5 competitors to find whitespace and parity gaps

Key Findings

  • Delivery delays were process failures, not engineering failures - requirements were ambiguous, PRDs were thin, and there was no proper sprint planning cadence
  • 80% of client complaints came from 3 feature areas - highly actionable
  • The highest-value feature (location-based lead attribution) was under-invested relative to its contract value
  • Two clients were near churn: addressable with targeted fixes in 6 weeks

What I Built & Fixed

1. Product Operating System

Introduced:

  • Weekly product review (roadmap, metrics, blockers)
  • Sprint planning with capacity tracking
  • PRD template with acceptance criteria, non-goals, and design decisions documented
  • Monthly client health review - RAG status per account Impact: Sprint commitment fulfilment improved from ~58% to ~83% in 3 months.

2. OKR-Based Prioritisation

Replaced HiPPO-driven prioritisation with a transparent OKR framework:

  • Company OKRs → Product OKRs → Feature priorities
  • Scoring matrix: revenue impact × effort × strategic fit
  • Roadmap published and shared with all clients quarterly This reduced “urgent requests from client X” noise significantly and gave engineering team clarity.

3. Location Attribution Feature (Flagship)

The highest-impact product investment: rebuilt the location-based lead attribution system. Problem: Retail and automotive clients needed to attribute offline store visits to online marketing campaigns. The existing system had ~35% data accuracy. Solution:

  • Rebuilt the attribution pipeline with improved geofencing logic and dwell time filters
  • Added campaign-level attribution dashboards for marketing teams
  • Integrated with 3 major ad platforms via API Result: Attribution accuracy improved to ~78%. Two at-risk accounts renewed. Became a key demo feature for new client acquisition.

4. Self-Serve Analytics Layer

Problem: Clients depended on AECC’s team for every report. High support load. Solution: Built a self-serve reporting module with configurable date ranges, metrics, and export options. Result: Support tickets related to reporting dropped by ~45%. Client satisfaction improved.

Results

MetricBeforeAfter (12 months)
Sprint commitment fulfilment~58%~83%
At-risk accounts30
Revenue growthBaseline~30% growth
Support tickets (reporting)High-45%
Location attribution accuracy~35%~78%

What I Learned

  • The best PM investment in a broken team is process - before features, fix how you decide and deliver
  • Shared roadmap visibility with clients changed the relationship from reactive vendor to strategic partner
  • Churn prevention is the fastest path to revenue - retaining at-risk accounts contributed more to the ~30% growth figure than any single new feature
Explore more work