What a launch is supposed to change
Launch reporting often lists everything that moved: page views, email opens, social mentions, press coverage, sign-ups. The list is long, every number is up, and nobody can say whether the launch worked. The problem is that the measures were collected after the launch rather than chosen before it, so there is no standard to judge them against.
The fix starts in the plan. The launch planning guide asks for goals with a KPIA metric tied directly to an objective. and a target for each, and those goals define success. A launch meant to drive adoption among existing customers succeeds if adoption rises, whatever happens to press coverage. A launch meant to open a new segment succeeds if pipeline from that segment appears, whatever the email open rate. Choosing the measures before launch also stops the team from searching the data afterwards for something that looks good.
It helps to think of a launch's effect as a chain. The launch reaches an audience, some of the audience pay attention, some of those take a first action, some of those adopt the capability, and the adoption shows up in retention, expansion or new business. Each link is a measure, and the links nearer the start move faster and are easier to attribute, while those nearer the end matter more and are harder to attribute. The general discipline of setting objectives and choosing measures is covered in the objectives and KPIs guide; this guide applies it to launches, and the product launches hub shows where measurement sits in the wider launch process.
Launch KPIs by tier
The measures should scale with the tier, just as the effort does. Measuring a minor release against pipeline targets wastes analysis; measuring a major launch only on traffic misses the point.
| Tier | Primary question | Leading measures | Outcome measures | Review depth |
|---|---|---|---|---|
| Major launch | Did it change the business: new segment, new revenue, better win rate? | Reach in the target audience, launch page traffic, demo or trial requests, sales conversations started | Qualified pipeline, win rate against named competitors, adoption among target customers, revenue influenced | Full review with product, sales and leadership |
| Feature launch | Did the customers who need it start using it? | Email and in-product message engagement, visits to help content, first use | Activation and repeat use, effect on retention or expansion where measurable | Short review within the product marketing team |
| Minor release | Did it reach the users it helps without causing confusion? | Changelog and help views, support contacts on the topic | Use of the improvement where tracked | A line in a monthly report |
| Silent release | Not announced | None beyond product analytics | Product team's own measures | None from marketing |
A measure tree makes the connection between a launch goal and its leading measures explicit. The top of the tree is the outcome, such as qualified pipeline from the new segment; beneath it sit the drivers, such as demo requests and the rate at which they qualify; and beneath those sit the launch activities that should move the drivers. Building one for each major launch shows which measures to watch and which are noise. The mechanics of building and maintaining trees belong to the KPI trees guide in the analytics root, and the widget below shows the shape.
Launch KPI tree
The AARRR model is another useful frame for launches of software products: acquisition, activation, retention, referral and revenue. It maps neatly onto a launch's chain of effects and reminds the team that acquisition is only the first stage.
Leading signals: awareness, traffic, demand
Leading signals are the measures that move first. They do not prove success, but they show early whether the launch is reaching people and whether the message is landing.
Awareness shows in reach within the target audience, branded search for the product or the new capability, mentions in social and press, and attendance at launch events or webinars. The question is not how large these numbers are but whether they come from the intended audience. A launch aimed at finance teams that gets wide attention among developers has reached the wrong people. Where the launch runs search ads, the guide to paid search for a product launch shows how weekly impressions on the new product name can be read as an early count of that branded curiosity.
Traffic and engagement show in visits to the launch page, the time spent there and the click-through (CTRShare of impressions that result in a click.) from launch emails, adverts and social posts. Comparing click-through by message and by audience shows which framing works, which feeds back into the messaging hierarchy.
Demand shows in actions with commercial intent: trial sign-ups, demo requests, contact-sales forms, upgrade requests and sales conversations where the launch is mentioned. The conversion rate (CVRShare of visits or clicks that convert.) from launch page visit to one of these actions is often the single most useful leading measure, because it combines whether the right people arrived with whether the message persuaded them. A conversion rate calculator helps compare it against the page's normal rate.
Leading signals should be read against a baseline. The same page, email list or audience before launch, or a comparable launch in the past, gives a reference point. Without one, a rise may simply be normal variation or seasonality. The control charts framework shows how to measure a metric's normal range from its own history, so a launch effect can be told apart from ordinary movement.
Adoption, activation and usage
For launches of new capabilities to existing customers, adoption is usually the measure that matters most. A feature nobody uses has not succeeded, however well it was announced.
Adoption has stages. Discovery is the customer becoming aware the capability exists, which the launch emails and in-product messages are meant to cause. First use is trying it once. Activation is reaching the point where the customer gets value from it, which needs a definition specific to the capability, such as completing a first report or connecting a first integration. Habitual use is returning to it repeatedly. Each stage can be measured as a share of the eligible customers, and the drop between stages shows where help is needed. When a feature is switched on for customers in stages, only those with access count as eligible, and the guide to launching a feature that rolls out gradually reads each rollout stage as its own group.
The usual pattern is a strong discovery stage followed by a weak activation stage. Customers hear about the feature, try it, and stop before it delivers value because the set-up is unclear or the benefit is not immediate. That is not a marketing failure alone, but marketing can help: onboarding emails that walk customers through activation, help content placed where they get stuck, and customer success outreach to the accounts with most to gain.
Usage data comes from product analytics, which product management usually owns. Agreeing before launch which events will be tracked, and who will report them, avoids the common situation where the launch goes out and nobody instrumented the feature.
Pipeline and revenue influence
For major launches, the business question is whether the launch produced or accelerated revenue. Answering it honestly is harder than it looks, and overclaiming does more damage to product marketing's credibility than reporting a modest result.
Three approaches are common, and the stronger reviews use more than one. Sourced pipeline counts opportunities created from launch activity, such as demo requests from the launch page. It is clear but narrow, since many buyers who see a launch arrive through other routes. Influenced pipeline counts opportunities where the launched capability was discussed or where contacts engaged with launch content. It is broader but generous, since engagement does not prove the launch made the difference. Comparison looks at win ratesShare of closed sales opportunities that were won., deal sizes or sales cycle lengthAverage time from opportunity creation to a closed deal. for deals involving the new capability against deals that did not, or before and after launch. It is the closest to evidence of effect but needs enough deals to be meaningful.
Win-loss notes add the buyer's own account: did the launched capability appear in the reasons for winning, or did a competitor's response appear in the reasons for losing? Sales leadership often trusts these more than attribution reports.
Return on investment (ROIProfit relative to cost.) for a launch is worth estimating for major launches, as long as the method is stated and the uncertainty acknowledged. Present influence as influence, sourced as sourced, and the comparison with its caveats. The monthly marketing report template has a suitable place for launch results alongside the rest of marketing's reporting, so that launches are judged in the same frame as other activity rather than in a separate celebratory document.
The post-launch review
The post-launch review closes a launch and improves the next. The Product Marketing Alliance treats it as a standard step in a mature launch process, and the format can be simple.
Hold it once the measures in the plan have had time to move: a few weeks for most feature launches, longer for major ones, with a final commercial read after a sales cycle. Invite the workstream owners. Cover four questions in order. What were the goals, and did we meet them? What worked, with evidence? What did not, with evidence? What will we change in the process for next time? Where the launch ran paid campaigns, the campaign wrap-up checklist confirms they have all stopped and their numbers are final before the review reads them. The campaign post-mortem template sets these questions out as a table, with a row for root causes and owners for each change.
The fourth question is the one that matters, and it should produce changes to the launch template, the tier criteria or the bill of materials rather than a list of good intentions. If sales was briefed too late, the template moves the briefing earlier. If adoption stalled at activation, the bill of materials gains an onboarding sequence for feature launches. If the tier was wrong, the criteria are adjusted. A post-launch reviewThe meeting that closes a launch once its measures have had time to move: goals against results, what worked, what did not, and process changes for the next. that leaves the template, the criteria and the bill of materials unchanged has closed the launch without improving the next one.
Write the review down, keep it short and store it with the plan. Over a year, the collection of reviews becomes the best guide to what launches the company can expect from each tier and each channel, and it gives the next round of planning realistic targets rather than hopeful ones. Reviews should look at the whole FunnelModel of the customer journey stages. the launch was meant to move, not just the stage that went best.
An expense-management company reviews a feature launch
Pellworth, an expense-management platform, launched receipt-matching for corporate cards as a feature launch, with a goal of 30% of eligible customers activating within eight weeks. The launch email reached 2,400 customer admins with a click-through rate of 11%, and 52% of eligible accounts tried the feature at least once. But only 18% activated, defined as matching a full month's card transactions. Product analytics showed most accounts stopped at the bank-feed connection step. The review added a guided set-up email sequence and a help video to the bill of materials, and customer success contacted the 150 largest accounts. Four weeks later activation reached 31%, and the feature-launch template now includes an activation sequence by default.
The paid media side of the launch scorecard, from burst reach to sustain conversion, is set out in media planning for a product launch, including the triggers for moving budget between phases.
If a launch has already missed its targets, diagnosing an underperforming product launch works back from the shortfall to tiering, positioning, sales readiness, skipped gates and late measures.
Common mistakes
- Choosing measures after launch. Goals set afterwards will always be met; set them in the plan.
- Counting attention as success. Traffic and mentions are leading signals, not outcomes.
- Measuring every tier the same way. Minor releases do not need pipeline analysis; major launches need more than traffic.
- No baseline. A rise without a comparison may be ordinary variation or seasonality.
- Ignoring activation. Customers who tried a feature once have not adopted it.
- Not instrumenting the feature. Agree tracked events with product before launch, not after.
- Overclaiming revenue. Influenced pipeline presented as sourced revenue damages credibility when finance checks it.
- Reviews that produce no changes. A review that does not change the template has not finished.
- Reporting launches separately. Put launch results in the regular report so they are judged alongside everything else.
