According to some publishers, programmatic is just a way for agencies to commoditize ad inventory and make sales teams irrelevant. The truth is, software has mostly replaced travel agents, and it will replace some ad salespeople and media buyers too. The benefits to advertisers are too compelling for this not to happen. Amex has already announced it is going 100% programmatic, and others will follow.
So hoping it will just go away if you do nothing is not a healthy option.
If you sell display sponsorships with an effective CPM of $1,500 and agencies aren't pushing you to demonstrate ROI, then go ahead and keep milking it. But start working on a plan B now, because as programmatic expands, you'll have fewer places to hideâ€“whether you choose to participate or not.
Many publishers still equate programmatic with low-quality remnant inventory ad networks notorious for "black box" programs that are rife with fraud. That is not the reality of programmatic now. Amex isn't going to buy crap audiences. They want the most efficient way to reach great audiences. Programmatic is starting to deliver that, and it's getting more efficient as competition and participation grow.
If you're already delivering performance-based campaigns and your audience is valuable and engaged, programmatic could theoretically deliver higher CPMs than you're getting now. Programmatic will almost certainly improve sell-through, meaning fewer unsold impressions every month. Forbes is generating 30 percent of digital revenue from programmatic selling already.
A lot of the fear and confusion comes from the singularity-like speed with which the industry is evolving. That pace will only accelerate. If you are delivering demonstrable value to your advertisers now, it's time to get working on your programmatic strategy.
A group of mobile tech companies participating in something called the Mobile Deeplinking Project made an announcement yesterday that you might have passed over. Have you ever opened a LinkedIn Groups update email on your phone and wondered (for the 500th time) why it couldn't just launch the app? Instead it drops you at the login screen in your browser, helpfully suggesting you download LinkedIn's fabulous app. That is the very definition of throwing away vast, profitable amounts of user engagement. How many people are really going to tap-tap-tap their credentials into the teeny browser (again) or manually launch the app and try in vain to find that discussion that first caught their eye? Nearly nobody, that's who.With mobile deep linking, tapping on that interesting discussion thread in your email would launch the LinkedIn app and deliver the exact page you were expecting, i.e. 'deep linking' you to the otherwise impossible to find page inside LinkedIn.So what does that mean for publishers who now see a path to making their apps a viable place to sell ads? If your 'app' is a simple digital replica of your print product, none of this will matter. If you have a useful app that extends your brand, then read on.Let's say you are like CNET and have well-regarded product reviews that millions rely on to make good decisions when purchasing electronics. If you put all those reviews into a well-designed app and promoted it well, you could get lots of people to download it. Then the app will most likely never get opened again, lost among dozens of other unused apps on millions of phone screens.However, if CNET sent out a weekly newsletter that leveraged deep linking, suddenly their app starts getting used, perhaps promoted to users' home screens and regularly updated. Otherwise, all that effort to develop the app will continue to show a crappy ROI.Â It is this new set of standards and best practices that has the potential to put the native app versus mobile web debate to rest. Now apps can be just another, slicker part of the web instead of these islands of awesomeness that are forgotten because they're too much trouble to get to. Publishers will be able to point to engagement data instead of downloads when pitching to advertisers.These standards are open source and therefore a moving target, so there will be some hiccups. But it's all laid out pretty well, complete with code libraries for your developers to leverage. If you have a great idea for an app that you've never built because you couldn't see how to get enough users to matter, this may be the time to revisit that decision.
Customer and audience frustrations are fertile ground for new revenue opportunities. Sometimes it's easy to map those frustrations to a new data product or service, and sometimes you hit a brick wall. When that happens, it helps to get outside your comfort zone to find solutions.1. Generate Data From ScratchHere's a real example: A data publisher supplies data about a certain commodity to the financial industry. They identified a need for real-time data about shipments of this commodity into and out of U.S. ports, but the buyers and sellers closely guarded the information and would not sell it. Undeterred, the publisher set up webcams at all major U.S. ports and used some simple software to assess shipment volumes based on how high in the water the ships rode. That company has created a proprietary data product that is sold at a premium to commodities traders who use this data daily. This "brute force" approach might seem a bit unrealistic for most of us, but it shouldn't. There is still plenty of opportunity to build new data products by going out and grabbing data that nobody has bothered to get yet, and then presenting it in a way that's immediately useful to someone.2. Create MashupsThere is also revenue to be mined from simply packaging or combining data to make it more useful or compelling. If you want to capture some of this revenue, you should do it quickly, because there are few barriers to competition. Someone is going to beat you to it or the government is already doing it well enough that it's not worth trying to monetize. Case in point: How do I find an unbiased, detailed evaluation of rehab facilities near my elderly parents' home? There are several commercial operations that have tried to make a go of this, but this Medicare site is better than any of them. That's because the ratings data is generated directly from government inspections of the facilities, and even the government is now figuring out how to present data usefully to consumers.A bigger example of this kind of disintermediation is how Google has obliterated traditional b-to-b directories and buyers guides. There is a way to transform some of these into data products someone is willing to pay for, but it's an uphill slog.3. Leverage Your Existing DataDo you do a "Top 10" or "Top 100" projects, products or technologies? Many awards programs can support a data product spin-off. Does your editorial coverage regularly include industry statistics and information your readers see as competitive intelligence? If so, and assuming you don't cover financial services, energy or pharma, there is likely an opportunity for you to create a data product, and to explicitly blend that with your editorial coverage. Done correctly, editorial can be the marketing funnel for a data product or services. Skift CEO Rafat Ali has some interesting thoughts on this concept. Wardsauto.com is a good example of this paradigm.Data product opportunities don't last forever, and those willing to roll up their sleeves and map the way forward are far more likely to capture the bulk of the profit potential.Â
If you are a media company, your top-selling product for 2017 probably hasn't been invented yet. This was one of the more headline-worthy findings of JEGI's 2014 Media Growth Study (pdf). You don't have to believe in Kurzweil's Singularity to see that the rate of change in information technology is accelerating. The shrinking time-to-adoption curves in this plot illustrate that technologies go from invention to mass adoption more quickly every year.
It's easy to be overwhelmed by the rate of change, and get caught up in chasing the latest technology as the savior of your business. But when new technology goes mass-market so quickly, the competitive advantage it confers to early adopters doesn't last very long. Soon everybody has it and it's just table stakes. All industries are in the process of being disrupted by information technology, even if they don't all see it yet. I believe the only way to maintain competitive advantage is to be better and faster than your competitors at translating customer needs into new business models. The promise of predictive analytics based on Big Data may allow deep-pocketed companies to build defensible market positions for themselves for a time, but automation and optimization can only go so far. We aren't yet at the point where algorithms can make the leap from vaguely-expressed customer frustration to ground-breaking new product concept. Where do you start? In the JEGI study, 31 percent of executives said they had a structured innovation program for generating new product ideas, while 71 percent said launching new products or services would be a top growth driver over the next two years. That's a disconnect you can take advantage of. Create and embrace a structure for innovation in your company and you'll innovate faster and more effectively. What does that structure look like? Sounds like a great topic for another post.
Is your website converting visits to revenue efficiently enough? Getting people to fill out a registration form or click through on an offer is difficult, but getting it right is critical in the battle for revenue and market share. The most effective thing you can do is to learn directly from your users how to lay out your site and what to offer. The benefits of this approach are well-documented, but the example that resonates with most of us is the ACA or "Obamacare" website. This site was built the way many publisher sites are, without doing any usability testing up front. We all know how that turned out.Let's review the typical approaches taken by publishers:Focus Group of OneUse internal staff as "stand-ins" for your users. Editors know their readers, so this should work, right? It's popular because it's really easy. The dismal results of this approach are everywhere. Research Best PracticesStudy the latest best practices in user interface design, lean UX, persuasion architectures, and the many sources of insight about how to remove friction from your conversion funnels. This is worthwhile, and applying these lessons to your site will likely generate measureable improvement. However, there are problems. Which expert do you believe? How do the cumulative results of user behavior on many other sites compare to your users, your industry and your content? Best practices give you insights like "send your newsletter on Tuesday morning" or "make action buttons bigger." They can't be specific to your users because they represent aggregated, normalized data, not detailed feedback. Researching and applying best practices is a lot better than the focus group of one, but it's still a scattershot approach.A/B or Multivariate TestingTools like Adobe Target, Google Experiments or Optimizely can allow you to test copy, layout and graphical elements to see which combinations yield the best conversion rate. What this kind of testing can't reveal is why users like Page A more than Page B. You might get lucky, but this kind of testing is best used to optimize products that have already been designed with direct input from your users. Usability TestingUsability testing involves the user going though a number of tasks while giving constant feedback to the researcher. Everything the user does is captured in a screen recording, and in a video that captures the user's reactions at every stage. In this way researchers understand in detail how well users are able to use the website or app, what confuses or annoys them, and what they'd really like to see but don't. Then you go build according to what you learned, and before launch, you test it again to make sure you got it right.Think about the last major brand site that left you frustrated. For me it was trying to add minutes to my kids' AT&T prepaid phones. An hour of bouncing between website and phone representatives finally finished what should have taken a minute of navigating and clicking. I can guarantee you that AT&T skipped usability testing, with a very frustrating result for millions of customers like me. Do you think a poorly-designed subscribe process is any less frustrating for your users?Usability testing isn't cheap (figure $20,000 to start) because it involves a lot of time, resources and logistics. That price tag is dwarfed by the amounts publishers regularly spend on designs that confuse and confound users. Publishers who apply usability as part of their regular process would prefer that you continue to view it as something that can be safely ignored. If you want to make more than incremental improvements in your next site redesign, budgeting for usability testing is a smart move.