Why Snapchat might revoke your approved On-Demand Geofilter
Think you found a bargain to have a geofilter running near a popular event? Think again -- Snapchat's new system might promise more than it can deliver.
What Snapchat giveth, Snapchat can taketh away, in terms of its new On-Demand Geofilters, that is. Even if you have a geofilter approved, you might find it removed, as the new on-demand system works through growing pains.
Marketing Land discovered this the hard way. Two geofilters we had scheduled to run during the upcoming SXSW conference were suddenly revoked this week, 11 days after they were approved. Here’s one of the suspension notices:
The reason? As the notices said:
The Geofence that you have selected overlaps with too many current or previously scheduled On-Demand Geofilters and cannot be purchased at this time.
The “geofence” refers to the area where our filter would have shown, during the time we wanted it to appear. Snapchat limits any particular area to seven geofilters of any type at one time. The rejection suggests that our geofilters couldn’t run because the limit had already been reached.
The problem with this explanation is that our geofilters should never have been approved in the first place, if there had already been too many other filters in place. Instead, they were approved, and nearly two weeks passed before Snapchat retroactively decided there was a problem.
So what happened? Our assumption was that rather than there being too many other “On-Demand Geofilters,” the low-cost ones that anyone can buy, some larger advertisers wanted to run their own filters through more lucrative deals with Snapchat, meaning that we, and perhaps other small advertisers, got booted.
When geofences collide
It’s not hard to understand the problem for Snapchat. Consider our geofence for these filters:
We purchased an unusual-shaped area covering the big hallways of the Austin Convention Center, where many SXSW attendees gather, as well as some of the nearby busy streets.
Our strategy was clever for a little advertiser: Select a small but highly frequented area where you’re not having to pay much. Our cost to cover these areas was about $30 for half a day.
A bigger advertiser with deeper pockets probably takes a broader approach, wanting to buy out entire city blocks to cover more people. But with Snapchat’s seven-filter limit, it’s possible that smaller advertisers buying just a portion of these areas potentially blocks the bigger buy.
Snapchat: The area should have been blocked
As it turns out, our assumption wasn’t too far from the mark. Snapchat’s public relations team referred us over to the On-Demand Geofilters support group. We’d previously contacted support but hadn’t heard back. This time, we got an explanation:
We understand you’ve had On-Demand Geofilters suspended and we wanted to provide more information. Unfortunately these Geofilters were placed over an area that our team had previously blocked off. However, some of your submissions slipped through due to our error and we apologize.
As we’ve explained, this tool can be turned off for areas where Live Stories or Sponsored Chain Geofilters are live, and we may restrict high-traffic areas and events. We appreciate your patience and understanding as we work through this new product!
Effectively, we were bumped for larger advertisers — but it was done retroactively, because the team that reviews on-demand filters apparently didn’t know that there were other advertisers already scheduled.
As said, the on-demand system is new — only out for about two weeks now. No doubt it’ll improve. Ideally, those using the self-serve system could see areas that have already been purchased and approved or blocked off for larger advertisers, before they go through the process of placing their own orders.
It also wouldn’t be surprising to see if the system evolves even more to require higher minimum buys for larger areas where Snapchat expects demand. Our survey of about a week ago found relatively little pricing difference for event-oriented buys. Those types of bargains, as our experience shows, are less likely to be available than they may seem.