Marketing 2020 4 weeks

From a full day of spreadsheets to daily optimization

A 15-person agency was losing a day every week pulling ad numbers by hand — and still couldn't prove ROI to the clients it was billing.

Weekly → Daily Optimization cycle
1 day/wk → 0 Manual reporting
1 week MVP shipped

The problem

Hook & Ladder is a marketing agency in Calgary running performance campaigns for national brands.

Every week, campaign managers pulled numbers by hand from six ad platforms across the entire client roster — a full day, sometimes a day and a half. Nobody saw performance in aggregate, so campaigns were optimized platform by platform, on instinct, and reports described activity rather than outcomes.

When an account manager asked how a campaign was doing, nobody could answer without digging. There was no data analyst on staff. I’d been hired as a campaign manager.

What I did

  • Spotted the gap and made the case — and got turned down. Analytics felt like overhead, and Funnel.io and similar tools had already been ruled out on cost and on a learning curve nobody had time for
  • Built a prototype on my own time — Stitch and Tableau, free, about a week including research. It covered only a subset of the platforms we used, but was real enough to show what the agency had been missing
  • Put it in front of the people who’d use it — campaign managers, who told me which parts they’d actually use
  • Turned the demo into requirements through stakeholder interviews, surfacing a non-negotiable: it had to stay maintainable by non-technical staff once I was no longer hands-on
  • Chose a stack leadership would fund, landing the full build at a fraction of the off-the-shelf quotes — which turned the original no into a yes
  • Built the production pipeline — Supermetrics ingestion from all six platforms plus client CRM where shared, into a Google Sheets layer that joined the feeds for per-platform and cross-channel slicing for the first time
  • Designed two dashboards for two audiences — slice-and-dice for campaign managers, a fast read on trends and channel mix for account managers
Hook & Ladder reporting pipeline Six advertising platforms feed a daily Supermetrics ingestion into a Google Sheets transformation layer, which powers Looker Studio dashboards built for two audiences: campaign managers and account managers. DATA SOURCES Google Ads Facebook Ads LinkedIn Ads Google Analytics StackAdapt Snapchat Ads + client CRM data where shared Supermetrics daily ingestion Google Sheets clean, join, aggregate Looker Studio two dashboards Campaign mgrs slice & dice Account mgrs trends at a glance I built paid tool sources & consumers
Six ad platforms into one daily pipeline, ending in two dashboards built for two very different readers

Results

  • Optimization moved weekly → daily, on data that refreshed every 24 hours
  • A full day a week of manual pulling, gone
  • Client questions answered in minutes, not days
  • A churning wealth management client stayed — the dashboards showed them ROAS against lifetime value, budget by channel, and conversions tied to specific campaigns
  • A new paid engagement: auditing another agency’s campaign work, a service Hook & Ladder couldn’t sell before

After launch

Those first dashboards were Tableau — richer visuals, plus auto-generated written summaries account managers could scan instead of decoding a chart.

A month in, leadership pulled Tableau. At roughly $900 CAD a month, it was more than they wanted analytics to cost. The value wasn’t in question; the price was.

So I swapped the visualization layer for Looker Studio and left the rest of the pipeline alone. Free, it integrated cleanly with the Sheets layer and met the maintainability bar better than Tableau had. The loss was those auto-summaries — so campaign managers wrote short narrative reads alongside each report instead.

Less elegant. But it fit a budget the agency would defend, and it survived me leaving.