Maya’s AOS Moment – 2
The setback
Six weeks into the pilot, Play 4 hits a wall.
The 30-day “no promotion” rule generates internal resistance Maya did not anticipate. Her e-commerce team panics about a perceived revenue dip from the test cohort during the promotional pause. The merchandising team escalates to the COO. The COO calls Maya: can we resume promotion to that cohort, just for the festive week?
Maya negotiates a compromise. Utility content alongside reduced promotional frequency, not zero promotion. The Play 4 pilot continues but with weaker conditions than the design called for. At 90 days, the B– cohort decline rate is 4% better than control — statistically present, but well below the 15–20% Maya had hypothesised.
She documents the learning. Play 4 worked partially. The structural issue was not the play; it was the governance. AOS assumed the CMO had unilateral authority over how a customer cohort is treated for thirty days. In Maya’s organisation, that authority is shared with merchandising and e-commerce. The play succeeded to the extent her authority allowed; it underperformed where the authority was diluted.
She takes the learning to the CEO. Not as a complaint about cross-functional friction — as a structural finding. “The reason this play underperformed was governance, not design. If we want to run AOS properly, we need to clarify which cohort decisions belong to marketing and which to merchandising. Here is the cost of the ambiguity.”
The CEO agrees to a cross-functional review.
The impact at six months
Play 6 has held. Paid Repeat Leakage at six months is 31%, down from 42%. Owned Repeat Ratio is 69%, up from 58%. The CFO has recalculated contribution margin against the new bucket distribution; the picture is meaningfully better than the old dashboard had been showing.
Play 5 is partial. The Atrium step worked: 9% of the R1 cohort returned to active attention within 30 days, comparable to industry benchmarks for daily-engagement programmes. The Meridian step worked unevenly: of the 9% restored to B–, 31% converted to B within the next 90 days. Recovery Conversion Rate of 31% is below Maya’s hypothesis of 45% but well above what the paid-retargeting control achieved. This split teaches the team something important: Atrium restored attention, Meridian converted value. Counting the first step as full recovery would have overstated Alpha. AOS forced the team to count potential Alpha and realised Alpha separately.
Play 4 is inconclusive at a transaction level but produced the most valuable structural finding of the pilot: AOS works in proportion to the CMO’s cross-functional authority.
Net Alpha Generated: $740K incremental contribution profit over two quarters, measured against the agreed Beta baseline. Most of it from Play 6; some from Play 5. Maya now has a credible AOS Dashboard, a renegotiated agency contract paying against Alpha rather than against ROAS, and quarterly board reviews structured around the ten AOS metrics. She has briefed the CFO of her sister brand in the holding company.
The board discussion changes. Instead of asking only about CAC and ROAS, the CEO asks about Paid Repeat Leakage. The CFO asks whether the B– governance issue has been resolved. The marketplace team gets a new quarterly target: identity capture rate. The CRM team stops reporting only campaign revenue and starts reporting customer movement across the TAT.
What Maya learned
Three things, distilled.
First, the diagnostic was more valuable than any single play. The 42% Paid Repeat Leakage number — by itself — would have justified the whole exercise even if every play had failed. Knowing the size of the problem turned every other conversation from opinion into negotiation.
Second, the play that underperformed taught her more than the play that succeeded. Play 6 confirmed what the diagnostic had implied. Play 4 surfaced a governance problem that AOS had assumed away. Both findings were valuable; the second was actionable in a way the first was not.
Third, AOS is a framework that admits its weak points. The 10–15% NeoMarketing rung claim is provisional. Her own Play 5 results came in below hypothesis. She did not have to defend AOS as flawless to her board; she had to defend it as a discipline that produces measurable results and honest critiques. That was an easier defence than the one her previous frameworks had asked of her.
* * *
The CMO Takeaway
| AOS principle | What the CMO does on Monday |
| Every transaction has a tax | Build the seven-bucket revenue view |
| Every customer has a state | Build the TAT and track movement |
| Every discount is an economic choice | Add Offer Tax to CRM and paid performance |
| Every repeat paid sale may be leakage | Measure Paid Repeat Leakage monthly |
| Every recovery needs sequencing | Atrium restores attention; Meridian recovers value |
| Every outcome needs a baseline | Measure Alpha above Beta, not against zero |
The trilogy is now complete. The Tax-onomy essay showed how to see Tax and Time. The naming essay gave the system its umbrella: AOS. This CMO playbook shows how to run it. The work now moves from essay to audit, from audit to pilot, and from pilot to proof.
Maya did not run more campaigns. She ran a better operating system. The next CMO is reading the Tax-onomy essay this weekend. Their AOS audit starts Monday.
Buy New efficiently. Own Repeat completely. Recover before paying twice.
