LinkedIn ad reporting and optimisation

Every report depends on your conversion tracking. If the Insight Tag sits on the wrong pages, or conversions are not defined, every number is wrong. Check that Campaign Manager is set up correctly before you build a report.

Then define three conversions: a lead (form completion), a meeting booked, and an opportunity created in your CRM. The last two will usually come from your CRM, for example HubSpot, not from LinkedIn itself.

Make a sheet in Google Sheets or a dashboard in Looker Studio. Export from LinkedIn Ads once a week, one row per campaign and audience.

Use these columns:

• Spend.
• Impressions.
• Click-through rate.
• Cost per click.
• Leads.
• Cost per lead.
• Meetings booked.
• Opportunities created.

Add two calculated columns: cost per meeting and cost per opportunity. These are the two numbers that decide where money goes. Click-through rate tells you about the ad. Cost per opportunity tells you about the business.

Always read from the bottom of the funnel upwards. Start with cost per opportunity, then cost per meeting, then cost per lead, and only then click-through rate.

An ad with a low click-through rate and a low cost per opportunity is fine. An ad with a high click-through rate and no meetings is a problem. It attracts the wrong people, or the landing page does not match the promise.

LinkedIn data is slow and noisy. Pipeline takes weeks to show. As a rule of thumb, I would not judge an ad on fewer than 1,000 impressions, and I would not judge an audience on cost per opportunity until it has produced at least five or ten leads.

Leads and clicks show up in days. Meetings take one to three weeks. Opportunities can take longer. Give each campaign a fair window of three to four weeks before you cut it.

Use simple rules so you do not decide on mood.

• Take 20 percent of the budget from the audience with the worst cost per opportunity and give it to the best one.
• Change one thing per campaign per week: the audience, the creative or the offer. Never two.
• Pause an ad only when it has had enough data and sits well below your target for two weeks in a row.
• Write each change in a log, with the date and the reason.

Small moves keep the platform's learning stable. They also tell you what caused a change in results.

The report gives you questions. Why does one audience respond better? Which hook gets meetings? Take those questions to Analyse click-through by audience segment for creative differences, and to Run monthly creative refreshes for replacing tired ads.

• Optimising for cost per click. Cheap clicks from the wrong people cost more in the end.
• Changing several things at once, so you cannot tell what worked.
• Judging a campaign after three days.
• Reporting leads without checking whether they became meetings.
• Keeping the report in LinkedIn only and never connecting it to your CRM.

• You can say, in one sentence, which audience and which ad produce your cheapest opportunities.
• The weekly review takes thirty minutes or less and ends with one decision.
• Cost per opportunity falls over two or three months, not just cost per click.
• Your log shows each budget change and what it did. If you cannot explain why a number moved, the report is not doing its job.