Sigal Bareket is cofounder and co-CEO of Taptica
By Sigal Bareket So, you have decided to do the right thing and market your application. Fortunately, you live in a world where everything is trackable, measurable and transparent. You may be thinking, “I’ll try my luck and bid low. If I can get engaged users on this bid, it’s a win-win!” But what if I were to tell you that in the advertising world, more is actually less – meaning that the higher you bid, the less you pay for a booking or paying user? If you think about it, it really makes sense. The more you pay, the better chance you have to hit high-quality media pockets. But how high should you go? And can you really hit your volume targets when bidding only on the high-end traffic? The key here is to find your balance point. How much it is worth paying, per media source, to get the right amount of new users at the desired level or engagement. So how do you find this balance point? The answer is PMS – it is not what you are thinking. Here is what the acronym stands for: Post-install event Multi-level pricing Substantiate (achieve comparable results through trial and error) Post-install events Talk about a real game-changer. Ever since Has Offers and other tracking systems have allowed post-install attributions, mobile advertising has gotten a high boost. Now advertisers and demand-side platforms (DSPs) can actually follow up on the users that are being acquired, and track their post-install behavior. What can you track? • Is the user we bought using the application on the second day? • Is he or she buying virtual goods? In what amount? • Is he or she booking a flight, restaurant table or a hotel room? How often? Having all this information can help you be granular with your UA decision making. Multi-level pricing In a world of post-install event tracking, there is no one-bid-fits all. The question of how much it makes sense to pay per acquired user has many answers – sometimes as many answers as the number of media sources from whom you are buying. The best practice would be thus: 1. Start with a bid that seems competitive enough, and run it for a week 2. Get a report of user performance, based on post-install event, per media sources 3. Adjust the bid according to each media channel performance 4. Keep on experimenting on a weekly basis (or read the next section) Substantiate (trial and error) The thing that I personally like the most about this industry is the freedom to experiment. Think of yourself as a scientist, with the entire world as your laboratory. First, open up a parallel campaign that bids 50 percent more than your highest current bid to see if there is a new segment into which you can tap. Next, try mixing in some incent traffic while keeping your ROI to help generate volume. Finally, experiment with creatives, profiles, media channels or any wild idea you have. There is no downside – you can monitor performance and stop any testing that does not work. Sigal Bareket is San Francisco-based cofounder and co-CEO of Taptica, a global performance-based demand-side platform. Reach her at sigal.b@taptica.com.