It’s the reverse ferret to end all reverse ferrets. On Monday, the Sun newspaper scrapped the daily images of topless women on Page 3. This morning, they were back.

In the intervening days, NoMorePage3, the campaign started by Lucy-Ann Holmes, took the lion’s share of credit for the paper’s decision. It seemed a petition started by one woman on found victory where thirty years of campaigning by Clare Short, Harriet Harman, and Caroline Lucas had failed.

For observers of the charity world, this is a neat case study for the possibilities of online campaigning: in less than two years, a movement conceived and delivered primarily on the internet has accomplished what many seasoned campaigners for women’s rights felt was an impossible task.

Judging the success of a campaign like NoMorePage3 is helped by its straightforward clarity of purpose: to change a single publishing practice in a single newspaper. Some of the hottest takes on this story zero in on the drawbacks of this simplicity and perhaps the reason why victory was so short-lived—particularly that the cancellation of Page 3 does not mean the wider culture of sexism that enabled and sanctioned it has also been defeated.

Indeed, in most circumstances, totting up victory and defeat in a campaign is far more complex. A typical charity campaign is long-term, involves many tactics running in parallel, and changes in response to the unpredictable political, economic or social context. You might make progress in one area and lose ground in another.

NPC’s recent guidance on measuring campaign impact, Closing in on change, underlines this point:

 Linking campaigning and outcomes is complex; you may plan exactly the same activities for two different campaigns and get two completely different outcomes. What worked before might not work again.

For researchers this aspect of campaigning work presents a particular challenge—it is nearly impossible to identify an appropriate counterfactual (there is no version of Page 3 we haven’t campaigned against), or to fully account for the complexity of the environment we’re operating in.

Indeed, the question ‘did the campaign succeed?’ misses the point. Knowing the ultimate outcome is only the first step of understanding. What’s more important—and more interesting—is why this particular campaign briefly succeeded, for example, or why the legislative approach taken by Clare Short in 1987 failed.

Ultimately, it is the job of the evaluator to ask tricky questions: how much did our campaign contribute to the outcome? What other organisations and actors were campaigning for similar aims, and how influential were they? Were changes in the external environment—cultural, social, economic, political—more important than our own work? Do decision-makers themselves recognise the campaign’s role?

If asking such questions sounds like relentless killjoyery at a time for celebration, it’s because it normally is. (One description of an evaluator favoured by my colleague David Pritchard is someone who turns up after the battle is finished and shoots all the survivors).

What does understanding why a campaign succeeded tell us? Evaluation is critically important in preventing us from wasting our time and resources pursuing strategies that have no impact. It helps us to understand where our skills and experience can bring value to ensuring change is achieved.

Without this insight, campaigners will continue to make the same mistakes, and ferrets will continue their endless dance, darting in and out of the trousers.