Operations
Part of Build it in-house or brief an agency, SAAS marketing operations that hold up
When to run a SAAS marketing launch review on data and sources
SaaS marketing launch reviews: when to run them, which data sources to audit, and how to check claims, tracking and data handling before and after launch.
What to take away
- A launch review is a data audit, not a demo. It examines the numbers your campaign produced, then decides what to change.
- Google's Search Central documentation, “Verify Googlebot and other Google crawlers”, matters because bot traffic can distort reported conversion rates.
- Advertising claims should comply with the CAP Code (the UK Code of Non-broadcast Advertising and Direct & Promotional Marketing); the ASA's sanctions powers make checks before launch important.
- The ICO's “Security” guidance is relevant to personal-data handling; its data protection impact assessment guidance helps teams consider data protection risks.
- No hands-on testing was performed. This framework comes from published rules, vendor documentation and editorial judgement.
When to run a launch review and what it assesses
A launch review examines four things: the traffic generated, the claims made, the data collected, and the handover from build to steady state. It does not assess whether the product is good.
Run the review before a campaign goes live or its first paid impression is served. Re-run it when a material change affects the claims, targeting, tracking or data collected; review every significant launch, then repeat the review 30 days later once early data settles.
Four things a launch review assesses
- Traffic generated
- Claims made
- Data collected
- Handover from build to steady state
Start with traffic quality. Sessions from data centre ranges make your reported conversion rate unreliable. For crawl checks, follow Google Search Central's “Verify Googlebot and other Google crawlers” documentation; the same verification principle helps identify other automated traffic.
Audit the records behind reported performance using the sources your team actually relies on: GA4 reports for site and event activity, Google Search Console performance reports for search traffic, CRM exports for trial and customer records, and relevant ad platform reports for spend, impressions and conversions. Reconcile figures to source exports and record the date and filters used.
Next, claims. Every performance figure on a landing page or ad needs a source you can show. Check non-broadcast ads against the CAP Code (the UK Code of Non-broadcast Advertising and Direct & Promotional Marketing); the ASA can act against misleading advertising, so claims need an owner.
Third, data. A launch usually collects emails, trial signups or event data. Your review should confirm what was collected, where it sits, who can access it and how long it is retained. Record this in the launch data map.
Before and after: what changes when a review is done properly
The comparison below shows how sign-off differs when a team relies on impressions rather than verified evidence.
Launch sign-off before and after review
Before review
- Traffic source
- Total sessions reported
- Conversion rate
- 4.2% claimed
- Claims log
- None kept
- Data map
- Unknown
- Sign-off
- Marketing lead alone
After review
- Traffic source
- Bot traffic filtered
- Conversion rate
- 2.1% verified
- Claims log
- Source and owner
- Data map
- Systems and access listed
- Sign-off
- Marketing, data, legal
| Sign-off area | Impression-led launch | Verified-data launch |
|---|---|---|
| Traffic | Impressions accepted without checking automated traffic | Analytics exports checked; Googlebot requests verified and other automated traffic considered |
| Claims | Vendor figures accepted without a traceable source | Each landing-page and ad claim has a documented source |
| Data | Collection, location and access are unclear | Data collected, location, access and retention are recorded |
| Handover | No named owner or dated follow-up | The marketing operations lead owns sign-off; unresolved actions have an owner and date |
Where a number cannot be verified, label it as an estimate and say so in the review notes.
Separating vendor claims from evidence
Vendors will say their tool improves conversion. That is a claim, not evidence. Record it as a claim until your own data supports it.
Independent evidence looks different: a rule, a ruling, a published standard or your own analytics export. The distinction matters because launch decisions get made where the loudest claim wins.
Ask each vendor for the method behind any benchmark they quote. Where the method is not published, treat the number as a planning assumption you have not verified.
What counts as independent evidence
Published regulation, platform documentation, your own verified analytics, and signed-off test results from your campaigns. Each has a traceable source and a date.
What stays a vendor claim
Benchmarks from a sales deck, case study percentages without methodology, and any figure you cannot reproduce in your own account. Record these, but do not let them drive budget. Log each one with a date so the next review can test it against your own data.
Where compliance fits in the review
Advertising rules apply from the first paid impression. Check applicable ads against the CAP Code (the UK Code of Non-broadcast Advertising and Direct & Promotional Marketing). The ASA can investigate misleading advertising and require an ad to be withdrawn. A review that ignores this is incomplete.
Data protection is the second thread. The UK GDPR and Data Protection Act 2018 apply to personal-data processing. The ICO's “Security” guidance covers appropriate security measures; reviewers should confirm encryption, access control and retention are documented.
Keep the paperwork: consent records, the retention schedule and the date each check was run. Store these with the source exports and sign-off so the evidence is traceable.
Roles and sign-off
One person should own the review, usually the marketing operations lead, with named contributors for analytics, legal and product. The owner decides what counts as evidence.
Sign-off should be explicit. A launch is not reviewed until someone signs that traffic data is clean, claims are sourced and the data map is complete. Anything unresolved becomes a dated action.
Give the owner authority to delay a launch where traffic data is unverified or a claim has no source. A reviewer who can only advise will be overruled.
Common questions
How often should a launch review run?
Run a review before a campaign goes live, whenever a material change affects claims, targeting, tracking or collected data, and after every significant launch. Repeat it 30 days later once early data settles. A single review at launch misses patterns that appear once traffic matures.
Does a small team need a full review?
Yes, but scaled. One marketer can still check traffic quality, source claims and list collected data. The framework shrinks; the questions do not.
Can a software marketing agency run this review for US?
An agency can gather the evidence, but sign-off should sit with someone inside your business who owns the data and claims.
What if the review finds a problem after launch?
Log it with a date and an owner, correct the claim or the tracking, and note the fix. A review that only confirms good news is not a review.
