Launch and Learn – Driving Campaign Success with In-Flight Optimization

Dec 01

JonathanMobile Advertising , , , , No Comments

As advertisers devote an increasing share of their media budgets to mobile advertising, experimental campaigns are morphing into sustained efforts.  Even those brands just getting started in the channel will have heard the success stories and won’t be satisfied with ‘seeing what happens’.

Results matter and maximizing ROI or driving toward pre-defined, tangible goals places increased pressure on effective campaign management and optimization. Clear and measureable campaign objectives are the essential starting point. Acknowledging mobile as a powerful direct response channel and building in conversion points will improve targeting efforts.

However, simply setting up your campaign and letting it run its course denies the opportunity to leverage the rich data set that mobile advertising generates. Device, time, content, creative, destination and conversion metrics are among the points across which a campaign can be evaluated. Knowing how to interpret and translate that data into in-flight campaign optimization can mean the difference between a successful campaign and a failure.

Let’s get some assumptions out of the way:

  1. You’ve done your homework and have a good sense of which devices your target audiences are likely to be using (or you’re casting a wide net across mobile users).
  2. You have a mobile-friendly post-click destination. If you’re driving to a wired web destination go back to the drawing board.
  3. You have access to media and conversion performance data in the shape of publisher/network reporting and post click tracking using Google Analytics or a mobile analytics service.
  4. You’re a savvy enough planner – or work with one – so your media buy is leveraging properties with good content affinity to your message and your audience.
  5. You’re working with a publisher/network that will let you adjust campaign targeting in-flight.  (I’ve yet to work with one that won’t; it’s in their best interest.)

At the very least you’ll be seeing the top-line reach and response stats – impressions, clicks, and click through rates (CTRs). These are helpful in providing a baseline for measuring your optimization efforts.  Content, geographic, time and device-specific reporting are the next level of mineable data.

The second baseline measure will be conversions. If you’ve managed to get consumers to click on your ad, don’t simply direct them to a flat landing page. Instead, have some response or engagement points to take advantage of the click. With that in place, you’ll be able to generate conversion rates or engagement levels.

Now that you have your campaign set up and you’re collecting performance data, here’s a list of questions you need to ask yourself to identify opportunities that will enable performance improvement via ongoing optimization:

Media Metrics

  1. What are you CTRs by creative unit? Run multiple calls to action or creative treatments and adjust weight according to performance.
  2. What are you CTRs by time? Focus on day parting to deliver more impressions when your message has the greatest currency.
  3. What are you CTRs by device? Device metrics go a long way towards validating audience profile assumptions and ensuring relevant offers, content and design strategies.
  4. Which content or context (e.g. location) dimensions are performing best? Cross-reference those results with creative, time and device CTRs.
  5. Are your mobile search keyword groups fully exploiting aligned location or urgency searches? Search is a low cost, high conversion channel when deployed sensibly and managed properly.

Conversion Metrics

  1. What content is generating the most views on your landing pages? Remember, mobile content consumption will likely follow a different pattern from web browsing.
  2. Are there choke points in your conversion funnel? With multi-page interactions (e.g. registrations), you may notice drop off at certain points. Streamline the process to minimize friction and maximize familiar behaviour triggers.
  3. What are the conversion comparisons between different devices? Could your content or design strategy be adjusted to take advantage of device browser or feature capabilities?
  4. How do your conversions compare across geographies? What happens when you layer on device data to that metric? How about time of day metrics? Tweak your media to follow those patterns.
  5. If you’re attributing revenue to your conversions, are some sources more profitable than others? How are different offers translating to conversion rates?

While some of these might seem obvious and different conversion points (e.g. landing pages, downloads, coupons, commerce, etc.) have unique dimensions to address, the fulcrum for mobile advertising success is the effective use of the added data layers that mobile devices, context and design/content strategies offer.

Be Smart. Tip the scales in your favour by acting on in-campaign reporting.

NOTE: This post appears as part of Mobile Marketer’s Classic Guide to Mobile Advertising 2010 package. You can get the full guide here.

Possibly Related Posts:


Helping Advertisers Capitalize on Publisher Mobile Analytics

Aug 24

JonathanAdvertising, Media, Mobile Advertising , , , , , 6 Comments

Mobile will save publishing. Maybe. The iPad will save publishing. Maybe. Some yet to be imagined device will save publishing. Maybe. Let’s face it. We don’t know what will save publishing.

Or do we? Has the answer been in front of us the whole time? Maybe.

Advertisers will save publishing. That is a statement I will get behind. What will save publishing is giving advertisers compelling reasons to spend money. (Side note: by publishing I am referring to any media company that drives a significant portion of their revenue through advertising)

I don’t work for a content producer, but I have in the past.  I’m on the agency side and work with publishers on mobile advertising campaigns. What I’ve experienced has led me to two conclusions:

  1. Mobile Publishing monetization models are still being worked out. (This is not what I want to talk about today but it’s a huge issue in mPublishing. And the crux may be what advertisers need to bring to the table vs. what services and solutions publishers should or can offer.)
  2. Campaign analytics and reporting are still too shallow or are rarely packaged in a way that gets agencies – and by extension their clients – as excited as they should be.

Here’s my argument:

Publishers need to deliver targeting capabilities and usage analytics robust enough to allow media buyers to make intelligent campaign planning decisions. Otherwise, monetization attempts will mostly fail as marketers search for solid foundations for calculating customer targeting, engagement and acquisition ROI.

That sounds awfully grim. The good news is that the higher than web click through rates and the ‘innovation’ play will likely be sufficient to deliver enough revenue to keep everyone pretty happy – for now, that is.  So, before that, here is my take on merging analytics and targeting in a way that makes everyone look and feel good.

I’ve broken the analytics scope into four domains: Connectivity, Consumer, Content and Conversion. Each of these domains has multiple dimensions and for the purposes of this piece I’ll keep it pretty high level.

Domain # 1: Connectivity

This is one area where most publishers are doing a good job for the most part.  That may be largely to do with the approach many have taken in offering up OS-specific native applications. I’ve found it’s rare for a publisher not to be able to execute device targeting. Many can even target specific models within a family of devices (e. g. differentiate Blackberry 9700 from 8900). Increasingly, OS and device targeting should be the minimum expectation.

The second connectivity dimension is geo-targeting. Again, most publishers seem to grasp the inherent importance of location and context in mobile and have built-in capabilities to target location to some degree.

What’s critical here is the increasing granularity in OS, device and geo-targeting to enable messaging and offers that are relevant to device capabilities and user habits and preferences.  Check out Google’s location powered AdWord units.

Domain # 2: Consumer

You could make the case convincingly that location is a consumer domain but here the consumer domain also involves demographic and behavioural dimensions.  Publishers need to know who their audience is by age, gender, income, and so on. This is a stock media targeting request and it’s what advertisers expect. Admittedly, the personal nature of the device can make this information more difficult to gather as privacy hawks circle. But it’s not insurmountable.

The Weather Network in Canada asks for, but doesn’t require, age and sex information upon app activation. That’s a good start.  The mobile ad network JumpTap has created an ad preference manager where consumers can select the types of products and services they receive ads for – and that’s even better.

If location is the ‘where’ and demographic data the ‘who’, then the next ask is the ‘when’. Publishers need to be able to deliver sophisticated day parting options to allow advertisers to heavy up-spend at those times when consumption is heaviest or when their message will have the greatest currency.

Domain # 3: Content

This is basic. Publishers need to have visibility into what content is being consumed and to what degree. Beyond that, there has to be insight into how that differs by OS, device, demographics and geography.

Of course, there’s room for conventional reporting like total unique visitors and page views in aggregate. What’s more powerful, however, is knowing page views by content category, time per visit, frequency of user visits and duration of each visit, content viewed per visit, and content sharing. Each of these represents a potential targeting opportunity and the ability to deliver this data and wrap a story around the implications of these insights will make the case more compelling.

Domain # 4: Conversion

This final domain is a little harder for publishers to report on as they might lack the downstream visibility.  The CPM monetization model also lets a lot of publishers off the hook due to its diminished focus on performance.  The other challenge is the desire to manage advertiser expectations. Committing to a click through rate or conversion percentage is more risk than most publishers could stomach.  However, I’d urge more publishers to meet advertisers half way and offer CPC or CPA models to demonstrate more tangible returns from advertiser campaigns. This suggestion might be driven by a personal bias for mobile advertising as a very compelling direct response tactic than a pure brand play. We’ll see if the iAd changes that view.

The promise of being able to deliver micro-targeting reporting is the ability to have an analytics dashboard that tells when a 34 year old male using an Android device, clicked on an ad about Offer X, while viewing a specific piece of content and a certain time of day at a specific location and then moved through to conversion.

By and large, this is possible. There are privacy and permission issues that must be respected, especially with the location dimension but such nuance is very real. And to be fair, location is not always relevant. However, layering in some of the content dimensions I outlined would create a more powerful story.

The gap is what advertisers and their agencies seem to ask for and how it’s being packaged by publishers.  If publishers are serious about maximizing revenue from the mobile channel, they need to start offering reports at that level of detail, packaging it in a way that makes is clear, compelling and actionable without having to be asked for it or it being an exercise in pulling teeth.

It may sound like I’m being too tough on publishers here. However, I do believe most publishers are working hard to figure out the space and overcome the challenges. If anything, I have sterner words for agencies and advertisers that don’t ask for or mine for valuable data to optimize their buys or deliver device- and consumer- unfriendly post-click experiences.  I still see mobile ad campaigns that click through to a wired web experience or just offer banal product information without any clear or compelling follow-on call to action. It’s shocking!

My closing call to action: If information is power, then the power of the mobile channel is potentially unprecedented. The unique, personal and contextual dimensions of the device enable a granular picture of ad interaction and response. And it shouldn’t require a multi-million dollar budget (I’m looking at you iAd…). The data exists and we should use it wisely and in a way that benefits all parties – consumers, advertisers, agencies and publishers.

NOTE: As I’m in Canada, this is mostly directed at my fellow Canucks. Publishers and agencies in other parts of the world seem to be more on top of things (correct if that’s wrong…). But if anyone gets something out of this piece, I’m happy.

I’d also really welcome dialogue on the topic from publishers. I’ll admit to not having seen the full scope of every publisher’s mobile advertising and analytics offering. If you’re a publisher and are taking steps or have solutions to bridge these gaps, please share.

Possibly Related Posts: