Owner, trigger and scope
The campaign manager owns this SOP: they build the campaign, run the procedure and decide when to go live. The peer reviewer is a second practitioner who did not build the campaign and who checks it before launch. The reviewer can stop a launch; the campaign manager cannot skip the review.
The trigger is any new paid campaign, any duplicated campaign, any paused campaign being restarted, and any change large enough to need a fresh brief, such as a new market, offer or objective. It applies to search, social, video and display campaigns on any self-serve platform.
The scope runs from an approved brief to the end of the first day of delivery. Strategy, budget setting and audience planning happen before the trigger and live in the brief. Ongoing optimisation, including the decision to pause weak ads, happens after it and has its own procedure. This SOP is purely about getting a correct campaign live and confirming it is working as built.
Before you start: inputs and access
| Input | Owner | Notes |
|---|---|---|
| Approved campaign brief | Planner or client | Objective, audience, budget, dates and the target KPIA metric tied directly to an objective. |
| Approved creative and copy | Creative lead | Released through the creative approval process |
| Landing page URLs, live | Web publisher | Never launch to a staging address |
| Tagged links | Marketing operations | Issued through the link request process |
| Naming convention | Account lead | The house pattern for campaign, ad set and ad names |
| Conversion actions | Analytics owner | Confirmed as recording before launch |
You need edit access to the ad account, view access to billing and conversion settings, and access to the analytics property. If the brief is incomplete, the campaign brief guide describes what it should contain; send it back rather than filling the gaps yourself. If the Google Ads account itself is new, work through the new Google Ads account setup tutorial first, so ownership, access levels, links and conversion actions are in place before this procedure starts.
The procedure
Check the brief
Read the brief end to end and confirm it states an objective, an audience, a budget, start and end dates, a destination and a target CPACost of each conversion. or equivalent. If a target cost is given, check it is plausible against the budget using the CPA calculator. Raise any gap or contradiction with the brief's owner now. A launch built on an unclear brief is checked against nothing.
Build the campaign
Build the campaign in the platform following the account's agreed structure. The Meta campaign structure guide and the platform help centres explain how campaigns, ad sets or ad groups, and ads fit together; this step only applies those decisions. Set the objective, optimisation event, budget, schedule, targeting, exclusions, placements and creative exactly as the brief specifies, and leave the campaign paused or in draft. For a search campaign, a filled search campaign build sheet gives the peer reviewer a written record of every keyword, negative, ad and setting to compare with what was built.
Apply naming and tracking parameters
Name every level using the house pattern from the naming conventions guide, so reports and filters work from day one. Attach the UTMURL tags identifying traffic source. parameters issued for this campaign at the level your platform supports, and confirm auto-tagging or the platform's click identifier is on where the account relies on it. Never type parameters by hand when a tagged link has been issued.
Verify tracking end to end
Click each final URL, confirm it lands on the live page without a redirect chain and that the parameters survive. Submit a test conversion where you can, and check that it appears in the analytics property and in the platform's conversion diagnostics. If a conversion action shows as inactive or unverified, stop and involve the analytics owner; launching against broken tracking misleads automated bidding from the start.
Peer review against the launch checklist
Hand the campaign to the peer reviewer with the brief. The reviewer works through the campaign launch QA checklist with the campaign open, comparing what the platform shows with what the brief asks for. Every failed item goes back to the campaign manager to fix, and the reviewer re-checks it. Record the reviewer's name and the date of sign-off on the ticket.
Schedule and go live
Set the start time in the account's time zone, confirm the budget and end date one last time, and enable the campaign. Check that ads move into review and note when they are approved. If any ad is rejected, read the stated reason in the platform's policy centre, fix the asset through the creative approval process and resubmit rather than editing copy on the spot.
Watch the first day
Within the first few hours, and again before the end of the day, check that the campaign is delivering, that spend is pacing roughly in line with the daily budget, that impressions are reaching the intended geography and that conversions or test events are recording. Do not judge performance yet. Record what you saw in the launch log and hand over to whoever owns optimisation.
A duplicated campaign caught at peer review
Ferncastle Outdoor, an invented camping retailer, duplicates last spring's tent campaign for a new season with a budget of £4,000. The campaign manager updates the creative and dates but misses that the copy still excludes no past purchasers and points 2 of 6 ads to a URL redirected in the autumn. The peer reviewer catches both while running the launch checklist. After the fixes, the campaign goes live on a Monday at 9am; the first-day watch confirms delivery, 11 recorded conversions and spend at roughly 95% of the daily budget, and the campaign is handed over to the weekly optimisation review.
QA checks
Before the peer reviewer signs off, every one of these must pass:
- Objective and optimisation event match the goal in the brief.
- Budget, schedule, end date and time zone match the brief.
- Targeting and exclusions match the audience section, including existing-customer exclusions where the brief asks for them.
- Names follow the house convention at every level.
- Destinations are live, load on a phone and do not redirect.
- Tracking parameters are present and match the issued tagged links.
- Conversion actions are recording and a test event has been seen.
- Creative is the approved version, in the right formats and with no placeholder text.
- Launch checklist fully ticked, with any exceptions noted and accepted by name.
Hand-offs and records
At the end of the first day the campaign manager hands the campaign to whoever owns optimisation, often themselves in a different role, with the launch log, the brief and any open issues. Ad disapprovals that are still pending, any delivery concerns and anything the reviewer accepted as an exception go in that hand-off explicitly. If a problem on a campaign handed over this way is found days later, the diagnosis of live campaign problems found too late traces which of its six causes left the problem unseen and where to fix it.
Records kept from this SOP are the approved brief, the peer reviewer's sign-off on the ticket, the launch log with the go-live time and first-day observations, and the list of tagged links used. Store them with the campaign's other documents so that a later reviewer can see exactly what was launched and when. When the campaign is reported on at month end, the launch log is the first place to look for an explanation of an odd first week.
Common mistakes
Most launch errors are not knowledge gaps; they are skipped steps under time pressure.
- Treating a duplicate as already checked. Copies inherit old dates, URLs and exclusions. Run the full procedure.
- Self-review. The builder reviewing their own work misses the same things twice. Use a second person.
- Hand-typed parameters. Typos split traffic in reports. Use the issued tagged links.
- Launching with an inactive conversion action. Automated bidding learns from bad signals. Fix tracking first.
- Editing rejected ads in place. Quick fixes bypass approval. Send them back through creative approval.
- Judging performance on day one. Early data is noisy. The first-day watch checks function, not results.
- No recorded sign-off. Without a name and date, nobody can tell later whether review happened.
