Success is not one number
“Is the site successful?” has no single answer, because sites exist for different reasons. A page that attracts a modest number of visitors can be a great success for a local plumber if those visitors phone, and a failure for a publisher that depends on large readership. Before you measure anything, decide what success would look like for your site specifically.
A useful way to do this is to write a short statement of the form: “In six months, success means ___, which we will see as ___.” The first blank is the goal; the second is the observable evidence.
What success can mean for different sites
A local or service business. Success is usually inquiries, calls, bookings or visits to the premises. Search visibility matters mainly insofar as it brings in people from the right area. A moderate rise in the right kind of inquiries is better than a large rise in visits from far-off places that cannot use the service.
An online shop. Success is orders and the value of those orders, plus signs the shop is healthy, such as fewer abandoned checkouts. More visits that do not lead to purchases may signal a mismatch between what pages promise and what the shop offers.
A publication or blog. Success may be a growing and loyal readership, measured by return visits, newsletter sign-ups or time spent reading, with the caveat that these signals are imperfect.
An informational or support site. Success might mean visitors find an answer and leave satisfied, which can look like short visits. Here, falling support requests can be a better sign than long sessions.
A brand-new site. In the early months success is often about foundations: pages are indexed, the site loads well, and the first visits arrive. It takes time for search engines to discover and assess a new site, so early numbers are typically small.
The indicators that show these things are covered under performance indicators. This article is about the bigger question of how to decide whether the work is paying off.
Compare before and after, carefully
The simplest test of an improvement is comparing a period before a change with a period after it. This is useful but easy to get wrong. Follow these steps:
- Record the baseline. Capture the relevant numbers before the change, over enough time to see their usual range.
- Make one meaningful change at a time where you can. If you rewrite titles, rebuild navigation and switch hosting in one week, you will not know which action mattered.
- Write down the date and details. A change log is the simplest way to keep this straight.
- Wait long enough. Search engines need time to revisit pages and reassess them, and the delay varies. Judging after a few days is usually too soon. Several weeks is more realistic for many changes, and some effects take longer.
- Compare equivalent periods. If your business is seasonal, compare with the same weeks last year as well as with the weeks before the change.
- Look at the right pages. If you changed ten pages, look at those pages’ results as well as the site total.
Hypothetical example
Imagine a small site improves the titles and descriptions on its service pages. In the four weeks before, those pages received a hypothetical 400 clicks from search. In the four weeks after, they received 460. The difference is 60 clicks, or 15 percent of 400.
That looks encouraging, but there are questions to ask before crediting the change. Was there a seasonal rise? Did the number of searches go up? Did a competitor drop out? Compare against pages you did not change, and against last year. If unchanged pages rose by a similar amount, the titles may have had little to do with it. If only the edited pages rose, the case for the change is stronger, though still not proof.
Be careful about cause
Numbers change for many reasons: updates to how search engines work, seasonality, news events, competitors’ actions, and your own edits. Correlation between your change and a result is a reason to be hopeful, not a conclusion. The wise stance is “this is consistent with the change helping,” especially when the pattern holds up over time and across several pages.
Watch for results that look good but are not
- Traffic up, outcomes flat. More visits are not success if the extra visitors are not the audience you want.
- Rankings up for phrases nobody uses. Position matters only for terms that bring relevant people.
- Short-term spikes. One good week is not a trend. Reading changes over time is explained under metric interpretation.
- Moving the goalposts. Redefining success after the fact to match what happened makes the exercise meaningless.
What if nothing changed?
A flat result is information too. It might mean the change was not significant, that it hasn’t had time to take effect, or that something else is holding the site back. Return to your evaluation findings and consider the next-biggest issue. Treat measurement as a loop: define success, measure a baseline, change something, check the result, and learn from it. Nobody can promise a particular outcome, and honest measurement is how you find out what works for your site.