In the festive weeks a higher cost per lead is only the season's fault if the rise is in CPM, the price of reaching people, while the ad's click rate and the form's completion rate hold steady; if clicks or completions fell, something in the campaign changed, and the festival is covering for it.
Every year it starts in late September. Cost per lead creeps up on one client, then three, then most of them. By the week of Dhanteras (6 November this year, with Diwali on the 8th) the whole roster looks worse, and every client asks the same question: what went wrong?
Often nothing did. More brands advertise in these weeks than at any other time of the Indian year, they all bid for the same people, and the price of reaching anyone goes up. But "it's the festive season" is also the easiest thing to say about a campaign that has genuinely broken, and the season gives cover to problems that would be caught in a quiet month.
Split cost per lead into its three parts
Cost per lead is three numbers multiplied together, and the festival only moves one of them. Add these three columns in Ads Manager, for the last 7 days against the 7 days before the rise:
- CPM: what it costs to show the ad to 1,000 people. This is the auction. When more advertisers compete, this is what rises.
- Link click-through rate: how many of those people click. This is the ad and the audience. Competition for the auction doesn't change whether your ad interests someone.
- Form completion, or conversion rate: how many who opened the form submitted it. This is the offer and the form.
If CPM is up and the other two have held, the campaign is working exactly as before; it simply costs more to reach the same people. That is the festival. If the click rate or the completion rate has fallen, something about the ad, the audience or the form has changed, and it would be costing you leads in any month.
The festival raises the price of reaching people. It doesn't make your ad less interesting or your form harder to fill in.
Compare each client with itself, not with the roster
The festive rise is not the same size for every client. A jewellery brand is in the middle of the most crowded auction of its year; a coaching institute recruiting for a January batch is barely in it. A rise that is normal for one is a warning for the other.
So judge each client against its own numbers. If the account ran through last year's festive season, those weeks are the fairest comparison: is this October worse than last October, not worse than August? If it has no festive history, compare the three parts above, which hold up without one. What doesn't work is one rule for the whole roster, such as "CPL up 30% means pause", because it is too strict for the brands in the thick of the season and too loose for the ones outside it. We wrote more on this in there is no good cost per lead.
What to do in the festive weeks
- Agree the festive budget with each client before October, in writing. A client who expected higher costs judges the campaign on leads; a client who didn't judges it on the shock.
- Don't edit everything at once when costs rise. Each significant change sends a campaign back into Meta's learning phase, and in a crowded auction that costs more than usual. See every edit restarts the clock.
- Watch lead quality, not only cost. Festive traffic includes a lot of people browsing offers. Call a sample of leads every few days, and if cheap ones are coming from outside the area you target, check where the budget actually landed.
- For clients whose customers buy in this season, such as property, gold, cars and electronics, a higher cost per lead can still be the best return of the year. Judge them on bookings and site visits, not on the lead price.
- After Bhai Dooj, expect costs to ease over the following weeks. Don't scale a campaign just because its cost per lead fell when the auction emptied.