The problem this solves
Lifecycle programs are built one flow at a time and never seen whole. The welcome series was created at launch, the cart flow during a conversion push, the win-back after a churn scare - each sensible alone, none designed against the others. Two years later the account is an archaeology site: a customer can enter three flows in one week, timing choices encode assumptions nobody remembers making, and at least one automation is running that everyone forgot exists. Nobody can answer the basic question of what the program does to a given customer.
The revenue picture is equally blurry. Flows earn unevenly - a couple usually carry most of the value while others tick along earning little and costing attention - but without a periodic accounting, effort follows recency rather than return. The flow someone touched last quarter gets optimized again while a top earner runs on copy nobody has read since it shipped.
A quarterly audit is the fix everyone endorses and no one schedules, because reading an entire account with fresh eyes is days of work that competes with the quarter's deadlines. It is also, precisely, a reading-and-analysis mission: traverse everything, map the overlaps, rank the earners, and write findings a team can act on. The audit needs to land in two forms - a document with the full reasoning, and a summary where the team actually talks.
How the mission runs
- The whole account is inventoried. The agent traverses your Klaviyo account and catalogs every live flow and its structure: triggers, email sequences, delays, and exit conditions. The inventory alone often surprises - forgotten automations, dormant branches, a test flow still quietly triggering - and it becomes the audit's factual foundation.
- Overlaps and timing are mapped from the customer's seat. The analysis reconstructs what customers actually experience: where flows can stack into a crowded week, where a delay contradicts the behavior it follows, where an exit condition fails to fire and someone keeps receiving a flow they have outgrown. Program-level problems like these are invisible flow by flow.
- Revenue per flow is ranked honestly. Using Klaviyo's flow performance data, the audit ranks what each flow actually earns, separating the carriers from the passengers. The ranking is presented against the effort each flow seems to receive, which is where the reallocation argument - optimize this, retire that - writes itself.
- Findings become fixes, ordered by leverage. The audit closes with a prioritized fix list: overlaps to resolve, timing to correct, underperformers to rework or retire, and gaps where a missing flow is leaving an obvious moment unaddressed. Each fix carries its reasoning and expected effect, so the list can be executed without re-litigating the analysis.
- The full audit goes to Docs, the summary to Slack. The complete findings are written into a Google Doc - inventory, maps, rankings, fix list - as the durable reference. An executive summary posts to #growth: the three findings that matter and the recommended order of work, linked to the Doc, so the audit enters the team's conversation instead of a folder.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A complete quarterly accounting of your lifecycle program: a Google Doc holding the full audit - every flow inventoried, customer-experience overlaps mapped, revenue per flow ranked, and a prioritized fix list with reasoning - plus an executive summary in #growth that gives the team the three findings that matter and a link to the depth. The program becomes visible as a system, on a cadence.
Make it yours
- Diff against last quarter: once the audit runs on a cadence, each edition leads with what changed - fixes shipped, rankings moved, new overlaps introduced by new flows.
- Narrow the lens for a problem quarter: point the same traversal at one lifecycle stage, like everything post-purchase, and trade breadth for depth.
- Turn the fix list into missions: have each approved fix composed as its own scoped follow-up mission, so the audit's output becomes queued work rather than a static list.
Frequently asked questions
Does the audit change anything in my Klaviyo account?
No. The audit is read-only against your account: it inventories, analyzes, and reports. Every recommended fix is exactly that - a recommendation in the Doc, waiting on your decision. Acting on a fix is separate, deliberately scoped work, so an audit can never quietly modify a live program.
Who sees what - the Doc versus the Slack summary?
The Google Doc holds the full detail for whoever owns lifecycle. The #growth post is deliberately shorter: the three findings that matter, the recommended priorities, and the link. Two artifacts because the audiences differ - deep enough to execute from, short enough that leadership actually reads it.
How does it judge a flow as underperforming rather than just small?
Context does the judging: a flow is assessed against its reachable audience and its job, so a small-but-efficient flow reads differently from a large-audience flow earning little. Where Klaviyo's data cannot support a firm verdict, the audit says so and frames the finding as a question to investigate.