The decision. The context. The outcome.

Decision

A consent setting nobody noticed cost $750K and sent CAC up by $2M.

With Consequential Pulse event monitoring

Context

Chapter 1: a Shopify consent misconfiguration cut first-party data to the ad platforms. Chapter 2: the 90% traffic collapse it caused took three days to find.

Outcome

Baseline: 3 days to detect, found by manual analysis

A consent banner nobody accepted cost us three days and $750K.

The Beard Struggle

Decision

International remarketing was dead. Nobody knew why for weeks.

With Consequential signal and flow monitoring

Context

Improper phone number formatting broke every Klaviyo SMS flow outside the Americas, quietly costing $19K a month.

Outcome

Baseline: weeks undetected, root cause found in 1 day

Every international SMS silently failed. Nobody saw it for weeks.

Decision

A $40K/mo revenue leak, now caught the same day.

With Anomaly Hunter on a rolling baseline

Context

A theme update quietly broke a checkout step. Traffic looked normal, so nothing flagged it until the monthly close.

Outcome

Baseline: 4-week average detection lag on silent leaks

We were losing $40K a month and it never showed up in a dashboard.

Solawave

Decision

Creative refresh time halved by ranking ads on revenue, not clicks.

With Creative Performance and Audience Overlap agents

Context

Working out which ad actually drove sales took two to three days of pulling reports and cross-referencing Shopify orders.

Outcome

Baseline: 6-week refresh cycle, creative ranked on CTR

We were scaling the ads with the best clicks, not the best revenue.

Hiya Health

Decision

Creative strategy rebuilt around 90-day LTV, not day-one conversion.

With Attribution Reconciliation plus Creative Performance

Context

Every ad was judged on day-one conversion because that was the only number fast enough to act on. Cheap converts churned by month two.

Outcome

Baseline: creative judged on day-one conversion only

Our best day-one ad was our worst 90-day ad.

Half Price Drapes

Decision

From hearing about problems from customers to catching them in under an hour.

With Anomaly Hunter and Pulse Pixel in real time

Context

Conversion problems surfaced through support tickets, hours or days late, and every tool reported a slightly different number.

Outcome

Baseline: problems reported by customers first, hours to days

We used to learn about breakages from our own customers.

ThirdLove

Decision

Ending the blended-metrics argument with one number everyone trusts.

With Attribution Reconciliation and Ask-AI

Context

Meta claimed one attribution number, Klaviyo another, Shopify a third. Weekly meetings opened with a debate instead of a decision.

Outcome

Baseline: four tools, four different revenue numbers

Every Monday was an argument about whose number was right.

Decision

Three tools became one after outgrowing what a single dashboard could answer.

With The full agent portfolio on one dataset

Context

Triple Whale was not broken. It just stopped answering enough questions once the brand crossed $120M GMV.

Outcome

Baseline: three tools, none answering the why

The dashboard told us what moved. It never told us why.

Decision

A regional margin drop caught two weeks before the board review.

With Cross-Store Reporting with anomaly alerts

Context

Margin looked healthy nationally while one region quietly lost money. Every report rolled up to a single number.

Outcome

Baseline: regional margin reviewed quarterly

I would have found out the same day I had to explain it to the board.

Decision

Four storefronts run from one Slack channel.

With Cross-Store Reporting piped into Slack

Context

Four brands meant four dashboards and four sets of alerts. The CEO logged into each one just to know if anything needed attention.

Outcome

Baseline: four dashboards, four alert streams

I used to spend Monday mornings logging into four accounts.

Decision

The conversions were happening. The ad platforms never heard about them.

With Consequential Pixel server-side recovery

Context

Browser-side tracking dropped a quarter of purchase events after iOS changes. Server-side recovery put them back into the optimisation loop.

Outcome

Baseline: browser-only tracking, post-iOS match rates drifting

A quarter of our conversions were never reaching the ad platforms.

Decision

Attribution was solved. Decisions still were not.

With Consequential NorthStar

Context

The team already had multi-touch attribution. What they lacked was anything telling them which campaign to cut and which creative to scale.

Outcome

Baseline: attribution platform in place, decisions still manual

Attribution told us the split. It never told us what to do on Monday.

Decision

Retention flows only work if the data reaching them is complete.

With Consequential alongside the retention stack

Context

Recovered first-party signal was fed back into the email and SMS platform, making post-iOS lifecycle programmes 30% more effective.

Outcome

Baseline: retention flows firing on incomplete post-iOS profiles

The flows were fine. The data going into them was not.

Decision

A billing failure spike looked like churn until someone checked.

With Anomaly detection on subscription signals

Context

Subscription cancellations climbed in a monthly report. The cause was a payment retry configuration, not customer intent.

Outcome

Baseline: subscription health reviewed in a monthly cohort report

A payment failure spike looked like churn for three weeks.

Decision

Replatforming off Magento without losing the numbers.

With Consequential through the migration

Context

A platform migration is where tracking, feeds, and flows silently break. Continuous monitoring made the cutover observable instead of hopeful.

Outcome

Baseline: replatforming with no continuous view of what broke

Replatforming is where tracking quietly dies.

Decision

Warehouse-native intelligence at $250M+.

With Consequential on the warehouse

Context

At enterprise scale the truth lives in the warehouse. Consequential reads it natively and puts marketing decisions on the same numbers as finance.

Outcome

Baseline: warehouse reporting refreshed weekly, marketing separate

The warehouse had the truth. Marketing had a dashboard.

Illustrative decision

How a pet brand avoided a $106K BFCM sales shortfall.

With Creatives ranked by store-confirmed return

Context

Meta said the hero ad returned 4.1x. Store-confirmed sales said 1.9x. The team found out ten days before Black Friday and rebalanced the budget.

Illustrative outcome

Baseline: budget planned on Meta-reported return alone

Meta ranked that video first. The store ranked it third.

Illustrative decision

440 more BFCM customers on the same ad budget.

With A new-customer cost ceiling, set before the sale

Context

A supplements brand capped what a new customer could cost at $45, then reweighted its $162K Black Friday budget across channels.

Illustrative outcome

Baseline: each channel judged on its own return, no new-customer cost limit

Setting $45 before the sale changed where every dollar went.

Illustrative decision

A home goods brand reset its BFCM target by $144K, three weeks out.

With Founder and Marketer daily reports

Context

Three weeks of daily reports showed sales running 9% below last year, not 20% above. The team cut the target and the ad budget before the weekend.

Illustrative outcome

Baseline: plan built on last year plus 20%, compared against last month

The report showed minus 9% three weeks out, while we could still change the plan.