User Acquisition Manager
Michael Chisholm is the User Acquisition Manager at Skout, Inc. For the past two years, he has helped Skout grow its user network globally through direct response mobile advertising campaigns on a number of different platforms. He also makes a pretty mean margarita.
Because we have a global network, we’re actively promoting the Skout app in a really large number of countries. On the one hand, this gives us the opportunity to drive volume where there might be less competition from, say, the gaming companies that tend to drive up prices. However, it also means localizing in a lot of different languages, tailoring your campaigns to those audiences, and making sure that you’re getting a return on your spend in each of those markets.
Having this kind of global distribution, what are some of the biggest challenges you face today?
It means looking at a lot of data on a daily basis! We’re running install campaigns, not branding, and so we need to be very vigilant about install costs and conversion rates from a download to a completed signup, because that backs directly into our cost per user. We also need to make sure that we’re driving enough female installs in each geo. Even though dating is only one part of the Skout experience, it’s still important to have a good balance of males and females in the network. Ever been to a party where it’s twenty guys and two girls? Yeah! Even if you’re not looking for a date, the ratio changes the whole dynamic. So we’re following these metrics for each market with each partner. And, of course, we’re competing against the gaming apps, so it gets challenging to find traffic at our price point and to be able to scale those campaigns. We’re doing it, but it takes a little effort.
Our two most important metrics, after install costs, are conversion rates from install to a completed signup and the percentage of female installs we’re getting. If either of those are lower than the threshold we’ve worked out with our partners then we have to pause them. If we’re dumping too many guys into the network, it doesn’t matter how good the metrics further down the funnel might be because it’s doing more harm than good in the long run. In terms of optimization, unless you can target by gender, like with Facebook, it comes down to finding the right publishers and producing the right creatives that get the attention of the audience we want.
How do you typically work with networks to differentiate between the percentage of male to female users you are acquiring, and ensure that the creatives are optimized for each gender? Outside of Facebook and the other major platforms with gender registration data, what are some ways in which you identify pockets of female users that have a high propensity to download Skout?
For good or ill, it’s really up to the network to find the right publishers because, typically, we have such little visibility into where their inventory is coming from or where our ads are being placed. The biggest levers we have to pull are sub_publisher reports and creatives. We only allow partners to use creatives we’ve passed on to them and those are all produced in-house. With the sub_publisher reports, we can at least tell our partners where we’re either getting enough or too few female installs.
You mentioned earlier that managing creatives in-house was extremely important. Why is this the case?
Two main reasons, really. One, producing your own creatives allows you to control your brand image and messaging. Two, it gives you creative control. In the past, we’d allow ad networks to produce their own creatives and they’d come up with the most base stuff you can imagine. Usually either girls in bikinis or those ads that trick you into clicking on them because they look like you have unread messages or some kind of prize waiting for you. That’s not the brand image we want to have and girls in bikinis are just going to drive a ton of male installs, and we’re much more interested in getting female users.
Do you typically do post-backs of user sign-up gender selections (of course without PII), to verify that a certain network is driving the purported amount of female or male users that they claim they are?
No. We can see those stats on our end after a user signs up and then we’ll send a generalized report to our partners on percent females per channel.
What are some of the strategies you use when moving into a new market, to grow the two-sided marketplace of male/female users in a balanced way?
You look at the existing organic traffic and see if you’re monetizing those users at all. If so, then we’ll usually run a campaign on Facebook to see how well we can do attracting female users. We can run it in-house and shut it off once we have enough data. We’ll run both male and female targeted campaigns to see what the split is. If it’s 9 to 1, or worse — and, believe me, I’ve seen worse — we’ll know it doesn’t make sense to try to run campaigns there. There’s a bunch of countries where we can get all the thirteen cent male installs we want, but we’d be lucky to get half a dozen female installs. It’s crazy.
Is publisher/sub-source transparency important for you? If so, do you typically ask the network to share a hashed ID of the their publisher partner, or do some networks give you the actual publisher name in which they are serving impressions?
Sub-publisher visibility is really important for optimizing campaigns, getting rid of poor quality traffic, and detecting fraud. I’d love to get the actual names of the sub-partners and publishers that our partners use but I don’t think that’ll ever happen. At least, not with any regularity.
If it’s utilized properly, I think it could be an amazing tool. From what I’ve seen, so far, it still requires a lot of upfront work and research. We’re starting to see what programmatic can do and it’s very exciting. I’m also a huge believer in really good creative, though. I got my start in San Francisco working at a really big advertising agency back when TV, print and radio was all anyone did. The end of the Mad Men era of advertising. And I still believe in the power of a really good ad. So, while getting your ads in front of as many of the right type of users as possible is super important, you still need to entice them to click on it.
Data. Looking at lots and lots of data. I know what our baseline metrics are for incent and non-incent, video and banner ads, and such. If I see that installs are converting to users at half the normal rate, something is up. If you sell me 500 installs and I get zero users out of them, I’m not paying for it. You just have to be very vigilant because there are shady networks out there and they will try to sneak in all sorts of junk and hope you don’t notice.
Yeah, absolutely. We buy traffic to drive organic. Part of that involves buying enough to drive rank in the app stores. Being visible when someone is searching for new apps is incredibly important. And with every new user you acquire, you hope that they get a few of their friends to download it, as well.
Do you think that the ever-increasing CPI rates to acquire users is sustainable? Also, what do you see as the future of mobile user acquisition?
No, I think that prices will eventually level off. There will be some kind of course correction, or we’ll have the Dot Com 2.0 bubble burst, or something else will happen. The smaller fish will get out of the game completely, more inventory will free up, and prices will stabilize.
Skout is the leading global platform for meeting new people and expanding your social circle. Co-founded by Christian Wiklund and Niklas Lindstrom, the Skout community spans more than 180 countries, is available in 14 different languages and can be downloaded on iOS, Android and Windows devices. In 2014, Skout facilitated more than 500 million connections, and more than 10 million Journeys were taken through its Travel feature. The Andreessen Horowitz-backed company was founded in 2007 and is headquartered in San Francisco.