The 60-Minute Lifecycle Marketing Audit for a Team of One (With the Gap Table an Assistant Can Fill)
Thirty days into the job, the first marketing leader owes the founder a document that says three things: here is what we have, here is what is missing, and here is the order we will build it in. For lifecycle marketing, that document is an audit, and most of the audit templates you will find are written for a different company than yours.
They are tech-stack reviews. They ask whether you own a marketing automation platform, whether it is integrated with the CRM, whether you have a data warehouse. At a Series A company with a team of one to three, those questions have known answers and none of them are the problem. The problem is coverage: for each stage of the relationship, does something run, does it fire on the right event, does it stop when it should, and does anyone look at it before it goes. This post is that audit. It takes an hour, produces a gap table and a build order, and the coverage half can be filled by an assistant connected to your marketing system.
A coverage review, not a tech-stack review
The dedicated lifecycle audit content is thin. Lifecycle Mechanics’ audit checklist is the best of it and is explicit that it assumes your platform infrastructure already exists. General SaaS marketing audits from agencies such as PipelineRoad and Kalungi cover the whole function and give lifecycle a paragraph. Appcues’ explainer on lifecycle marketing asks the right question, what messaging exists at each stage and what the gap is, and never turns it into a worksheet.
Here is the worksheet. It has six columns, and the audit is the act of filling them in.
The hour, minute by minute
Minutes 0 to 10: list the stages. Write down the stages of your customer relationship as they actually happen, not as a template describes them. For a demo-led company that is demo requested, demo held, proposal, stalled, closed-won, kickoff, first value, renewal window. If you have not defined them, use the stage-and-trigger map for sales-led SaaS and skip the argument for now.
Minutes 10 to 30: inventory what runs. Every automation, sequence and workflow that sends to a prospect or customer. For each: what triggers it, what steps it has, when it stops, who approves it, and whether it has a quiet-hours window. Do not judge yet. Just list.
Minutes 30 to 45: fill the gap table. One row per stage. Mark each cell from the inventory. A stage with no row in the inventory is red. A stage with a row that is missing an exit, an owner or a window is amber. Everything else is green.
Minutes 45 to 55: mark the orphans. Walk the inventory against the five patterns below and note every hit in the last column.
Minutes 55 to 60: set the build order. Red rows first, ordered by the value of the stage; then amber fixes, which are usually five-minute edits.
The gap table
This is the template, with example rows filled in for a fictional company, Northwind, as an audit might find them.
| Stage | Automation exists? | Trigger: event or clock? | Exit rule | Approver | Quiet hours | Status and notes |
|---|---|---|---|---|---|---|
| Demo requested | Yes, 2 steps | Event: form submit | On meeting booked | None | No | Amber: no window; sends at 11pm |
| Demo held | No | Red | ||||
| Demo no-show | No | Red | ||||
| Proposal | Yes, 2 steps | Event: deal stage | None; runs to completion | None | Yes | Amber: no exit on reply; no owner |
| Stalled | No | Red | ||||
| Closed-won | Yes, 3 steps | Event: closed-won | On kickoff held | Founder on step 1 | Yes | Green |
| Kickoff | No | Red | ||||
| First value | No | Red | ||||
| Renewal window | Yes, 1 step | Clock: 30 days before end | None | None | Yes | Amber: clock where a date event exists; 30 days is late |
Nine stages, one green. That is a normal result for a company at this stage, and it is a useful one, because it turns “our lifecycle marketing is weak” into five specific things to build and three specific things to fix. The renewal row is the classic amber: it runs, so nobody worries about it, and it fires a month before the contract ends with no owner and no follow-up.
The five orphan patterns
Almost every amber cell in a small team’s audit is one of these.
- No owner. An automation nobody is named on gets edited by nobody, and its copy references an offer from two quarters ago. Every row needs a name, even if all the names are yours.
- No exit. A sequence that runs to completion regardless of what the contact does will, eventually, send a “just checking in” to someone who signed last week. Every automation exits on reply and on the event it was trying to cause.
- Clock where an event exists. “Send on day 3” is a fallback for when you cannot see what happened. If the CRM knows the meeting was held, trigger on that. The difference is the subject of drip campaigns vs nurture campaigns.
- Duplicate sends across a campaign and an automation. The proposal-stage automation sends a case study on day 14; the monthly campaign sends the same case study to the full list on the same Tuesday. Neither system knows about the other unless the calendar does. This is the one the audit most often misses, and it is the reason the content calendar mistakes post exists.
- No quiet hours. A trigger that fires at 11pm sends at 11pm. Every customer-facing automation gets a 7am to 9pm window in the contact’s timezone, without exception.
The build order
Red rows first, ordered by what the stage is worth. For a demo-led company that puts the no-show rescue and the stalled row at the top, because they recover deals that are otherwise lost to silence, followed by kickoff and first value, which decide whether the customer renews, and then demo held. The renewal row’s amber fix (move from a 30-day clock to a 90-day date event with an owner) is a five-minute edit and goes first among the ambers.
Then the audit is a slide: a nine-row table with colors, a five-item build list with dates, and a sentence about what the first green row is already doing. That is the day-30 document, and it is also the honest version of what to expect in the first six months: most of the work is filling red rows, and most of the reds are cheap.
Let the assistant fill the coverage half
Minutes 10 to 45 of the hour are reading. If your automations run in Marqeable, an assistant connected through its MCP server can do that reading and produce the gap table for you, because the connector exposes exactly the things the table asks about: every automation and its plan (trigger, steps, timing, exits, gates, window), the content calendar with scheduled sends, and semantic search over existing content. Marqeable’s server is at https://app.marqeable.com/mcp, OAuth 2.1 with your normal login, added in Claude under Settings > Connectors or in ChatGPT under Developer mode. Its writes are drafts only; there is no send, launch or publish tool.
You: Run get_automation_overview, then get_automation_plan for each automation, then read the calendar for the next 30 days. Here are my nine stages [pasted]. Fill in the gap table: for each stage, does an automation exist, is the trigger an event or a clock, what is the exit, who approves, is there a quiet-hours window.
Claude: Three automations found. Demo-request follow-up: event trigger on form submit, two steps, exits on meeting booked, no approver, no quiet-hours window. Proposal nurture: event trigger on deal stage, two steps, no exit rule, no approver, window set. Closed-won welcome: event trigger, three steps, exits on kickoff held, founder approval on step one, window set. Stages with no automation: demo held, demo no-show, stalled, kickoff, first value. Renewal: I found no automation, but the calendar shows a recurring “renewal reminder” campaign send 30 days before quarter end, which is a clock, not a per-contact date. [Gap table follows.]
You: Check for duplicates between the proposal nurture and the campaigns on the calendar.
Claude: The proposal nurture’s day-14 step sends the finance-team case study. The “Fall pilot” campaign has the same case study scheduled to the full list on the 18th. Any contact who entered proposal stage on the 4th receives both on the same day. I searched content and confirmed the two pieces share the same body.
You: Good. Draft automations for demo no-show and stalled first, from my map, with quiet hours and exits. Add a suggestion on the proposal nurture to exit on reply.
Claude: Two automation drafts created with the map’s timing, exits on rebook and on reply respectively, quiet hours on, approval gates on the SMS steps. I also drafted a revision of the proposal nurture that adds an exit on any reply from the account; the live version is unchanged until you approve. Links: [no-show draft] [stalled draft] [proposal revision].
The table took the assistant a minute, the duplicate it found is the one a human audit usually misses, and every fix it proposed is a draft waiting for a decision. The stage list, the build order and the approvals are still yours. That division, reading and drafting by the assistant, judgment and publishing by a person, is the whole method in building your lifecycle program with Claude or ChatGPT.
Run the audit quarterly, not once. The first one finds the red rows. The later ones find the ambers that crept back: the automation whose owner left, the campaign that now collides with a sequence, the offer that changed in step three of something nobody has opened since March. Strategic Flow reports that teams taking its email audit more than once see the score climb sharply between rounds; the pattern holds for lifecycle coverage too.
Frequently asked questions
What is a lifecycle marketing audit?
For a small B2B team it is a coverage review: for each stage of the customer relationship, does an automation exist, what triggers it, does it exit correctly, who approves it, and does it respect quiet hours. It is not a review of which tools you own. The output is a gap table and a build order.
How long should a lifecycle audit take?
About an hour for a team of one to three: ten minutes to list the stages, twenty to inventory what runs, fifteen to fill the gap table, ten to mark the orphan patterns, and five to set the build order. Longer audits are usually tech-stack reviews in disguise.
What are the most common lifecycle automation problems?
Five patterns account for most of them: automations with no owner, automations with no exit rule, time-based sequences where a real event trigger exists, duplicate sends where a campaign and an automation reach the same person in the same week, and sends outside working hours.
Can an AI assistant run the audit?
It can fill the coverage half. Connected to Marqeable, the assistant reads every automation and its plan, the content calendar and existing content, and produces the gap table for you. It can then draft an automation for each empty row. Deciding the stages and the build order stays with you, and every draft waits for your approval.
The bottom line
A lifecycle audit for a team of one is an hour with a nine-row table: list the stages, inventory what runs, mark each stage red, amber or green, name the orphan patterns, and set the build order. Most companies at this stage find one green row, which is fine, because the reds are cheap to fill and the ambers are five-minute fixes. Let an assistant do the reading and the drafting, keep the judgment and the publish button, and the day-30 document writes itself.
See the rows fill in: Marqeable’s agents build and run your campaigns, run behavior-triggered automations across text and email, answer and qualify every visitor with AI website chat, and tie revenue to the exact message with attribution.
Marqeable runs your campaigns, answers every visitor, text, and email in seconds, and turns them into booked jobs and meetings - even at 9pm on a Saturday. We’re in private beta with a small early cohort. Get early access
