The most damaging thing a business owner can do with marketing data is make major decisions based on two to four weeks of campaign performance. New campaigns are statistically volatile in their first month. Ad platforms are still learning audience signals. Seasonal patterns have not emerged. The early data tells you almost nothing reliable about long-term channel performance.
But the pressure to evaluate quickly is real. When you are spending $2,000 per month on advertising, month two feels late to wonder if it is working. Here is how to think about what data is useful at what point.
Week one to thirty: ignore the noise
In the first month of a new campaign, the only data point worth watching is whether leads are happening at all, not how many. A campaign that generates zero leads in the first 30 days has a structural problem worth investigating. A campaign that generates 5 leads in one week and 2 in the next has normal variance.
The tendency is to tweak the campaign based on early performance signals. Adding new keywords, changing ad copy, adjusting bids. Every change resets the learning period for the ad platform’s algorithm. Frequent early changes extend the noise period rather than shortening it. The better approach is to set the campaign up correctly and then resist the urge to optimize until there is enough data to see a real pattern.
Month two and three: the pattern emerges
At 60 to 90 days, with consistent campaign settings, the pattern becomes visible. Is the cost per lead trending down as the algorithm optimizes, or is it stable, or is it trending up toward a competitive ceiling? Is conversion rate stable or improving as landing page data accumulates? Is there a day-of-week or time-of-day pattern in lead quality?
These patterns inform the first meaningful optimization decisions. Which ad groups are significantly outperforming others? Which keywords are generating clicks but not conversions? Which audience segments are producing higher-quality leads? At 90 days with a real campaign budget, you have enough data to see these patterns clearly.
Month four through six: the channel judgment
By month six, you have enough data to make a genuine channel judgment: is this channel producing leads at an acceptable cost, and are those leads converting to revenue at a rate that justifies the investment?
The month-six view also reveals seasonal patterns that were invisible in the first quarter. A roofing campaign that looked mediocre in December might look strong in March. A dental campaign that performed well in January might slow in July. Understanding these patterns is essential for budget allocation decisions across the year.
Six months of data also reveals the lead quality dimension that click and lead data does not capture. If you have been consistently tracking lead source through to close, you now know whether the channel is producing leads that convert at the same rate as referrals, better, or worse. Cost per closed customer, not cost per lead, is the metric that determines channel ROI.
What to do with the six-month read
Channels that produce acceptable cost per closed customer at month six should scale. Channels that are producing leads but converting poorly need investigation: is the problem with the leads (wrong intent, wrong audience) or with the sales process that handles them? Channels that are producing neither leads nor converted customers should be cut or significantly restructured.
The business that makes these decisions based on six months of data rather than six weeks will make significantly better resource allocation decisions. The patience required is real. The better decisions are worth it.
If you want a framework for tracking the right data from day one of a new campaign, let’s build it before you launch.