Digital Tap Digital Tap
Explore Performance

Digital Tap  /  Methodology

Methodology

How Digital Tap
Evaluates Performance.

Six steps, applied the same way to every account. The sequence is deliberately ordinary — the value is in doing it consistently and in showing the working.

01

Collect

Gather available channel data.

  • Connect Google Ads, Meta, Google Analytics 4 and Google Search Console through platform APIs.
  • Confirm which metrics each account actually exposes before anything is reported.
  • Record the date range every figure was pulled for.
02

Normalize

Bring metrics into a comparable reporting structure.

  • Align naming so a campaign means the same thing across rooftops.
  • Separate platform conversion events from anything that could be read as a vehicle sale.
  • Flag partial periods — a month still in progress is never treated as a complete one.
03

Compare

Evaluate performance across time periods, campaigns, and channels.

  • Period over period, and year over year only where prior-year history exists.
  • Campaign against campaign inside an account, and rooftop against rooftop inside a group.
  • Channel against channel, so paid search and social are read together rather than separately.
04

Identify

Find meaningful changes and anomalies.

  • Separate movement that is large enough to act on from ordinary week-to-week noise.
  • Locate where a change originated — campaign, market, device, or search term.
  • Note what is missing as carefully as what moved.
05

Interpret

Add automotive context.

  • Inventory, incentives, OEM campaigns, geography, seasonality and store operations.
  • A CPL change during an incentive period means something different than the same change without one.
  • State the assumption when an interpretation depends on one.
06

Act

Use insights to guide optimization.

  • Turn each finding into a specific recommendation with an owner and an expected effect.
  • Re-measure the same metric over the same window after the change.
  • Report what did not work alongside what did.
Collect→ Normalize→ Compare→ Identify→ Interpret→ Act
What this method refuses to do. It does not report a figure without its period and source. It does not describe a platform conversion as a vehicle sale. It does not publish return on ad spend without connected revenue data. It does not draw a trend from a period that has not finished. And where a metric is unavailable, it says so rather than estimating one.
See It Applied Vendor Readiness