Mike Jacobs is chief services officer for iMarketing, an online marketing agency serving clients like Yahoo, Dow Jones, eMusic and AA RP. He is a consultant to several hedge funds and was previously co-founder and marketing director for Expression Engines/Bigfoot Interactive (sold to Epsilon in 2005). More information about iMarketing is available at www.imarketingltd.com.
PERFORM: How do you see Web 2.0 technologies fitting in with performance marketing?
JACOBS: It would probably be helpful to start with some definitions, because many people use these terms interchangeably. We define Web 2.0 as the next generation of online services, generally focused on collaboration. Social networking and wikis are prime examples, and older concepts like viral and lead generation are being included in the discussion as well.
Performance marketing is ROI-driven marketing. In theory, all marketing should be driven by ROI, but that isn’t the case right now. Performance marketing focuses on generating revenue to directly impact the bottom line. While brand-focused marketing can eventually help in driving revenue indirectly, performance marketing is generally executed with specific time frames and transactions in mind – sales, subscriptions, etc.
So far, Web 2.0 – particularly the version being sold by agencies – has had little impact on performance marketing. Web 2.0 creates avenues for connecting with consumers, but the commerce hasn’t yet followed as much as the hysteria has.
We’ve seen social networking sites selling tons of ad impressions without resulting in enough sales to justify the cost. When it is so easy for a consumer to create content – through a blog or a MySpace profile – the odds of that content actually being worthwhile are low. Even where page views are generated, you usually see low clickthrough rates and even lower conversion rates. There are some niche players who can use this kind of inventory – ringtone players being one example – but most large-scale advertisers are just not seeing the return.
This lack of performance is generally ignored for two reasons. First, marketers are too often getting caught up in the hype and not looking at the results. They are running campaigns without a clear view of goals and metrics. Second, numerous entrenched parties promote and profit from this confusion. Google, for instance, blurs the distinction between search and non-search media when selling to advertisers. A marketer may think he’s only advertising to users who searched on a particular keyword, but his ad could end up on a blog, on a MySpace page or even on a cybersquatter landing page. He may also pay the same price for these clicks – even though they are worth less – because AdWords makes it difficult to see or optimize the different traffic sources. Until marketers get smarter and demand accountability, these trends will continue.
PERFORM: We’re also seeing a lot more agencies out there offering to create buzz and develop viral campaigns. Does that fit in with your view of performance marketing?
JACOBS: Short term, going strictly by the numbers, no. Viral may never fit into performance marketing because it is nearly impossible to directly attribute results. Viral can enhance many campaigns and products, but it is not a marketing strategy unto itself.
I can’t tell you the number of potential clients who have come to us and said, “We don’t want to spend real money on marketing because our product is going to be viral.” But consumers are on to us. They know that we want them to sell our products for us. After years of getting hit by hundreds of “tell a friend” and other referral options – many making unfulfilled promises – consumers are weary of viral marketers. Furthermore, you need an established base of users at the outset. Viral growth is hard to start with just a handful of users.
“Buzz marketing” often confuses cause with effect. Many large interactive agencies offer “buzz measurement,” which isn’t marketing at all, let alone performance marketing. Talk comes from usage; not the other way around. Consumer feedback can be helpful, but measurement of the buzz often only measures the success of your acquisition campaign.
Some agencies will claim they can reach out to thousands of influencers on your behalf, in theory to drive activity. However, what they usually don’t tell you is that each of those same influencers got 100 similar Emails that day, often from the same agency. Even when we see a resulting product review or blog post, the results just don’t show. Look at referral sources for new customers. Track the referral rates from within the product itself. The numbers rarely match the story. Viral marketing can work, but what is being sold today as viral rarely delivers ROI.
PERFORM: That goes back to the first point, which was, people really don’t know what they’re talking about sometimes when it comes to Web 2.0.
JACOBS: Absolutely. Many marketers are simply acting out of fear. Web 2.0 isn’t well-defined, yet execs worry that they need to be doing something in the space, even if it means they are spending money without results. But that isn’t performance marketing. When the results justify the cost, only then should the money follow. The smart marketers will choose the channels that deliver, even if it means watching others get cut on the bleeding edge.
PERFORM: How about developments in the lead generation space? How do they impact performance marketing and affiliate marketing networks?
JACOBS: Lead generation is one of the core developments in performance marketing. It appears nearly risk-free on the surface, in that you only pay for results. But lead-gen for the most part works only for select categories of advertisers, like mortgage, education and debt consolidation. A close look at lead-gen companies will show that they are almost exclusively focused on these areas, and they have not really had a major impact on e-commerce sites.
There are also very significant risks that are not being taken into account. In reality, lead-gen reduces only CPA risk and creates additional risks like brand and lead quality. The value of leads for even the same offer can vary dramatically by source, yet lead quality rarely seems to factor into the equation. Even where folks are scrubbing lists to produce valid leads, it still doesn’t mean that those consumers actually have an interest in the product. Even valid leads can differ in quality, and incentive sites are a big part of the problem. If a consumer is told that she’ll get a free iPod when she signs up for The New York Times, she is much more likely than a non-incentive lead to cancel her subscription soon after. Leads from different sources have different values, and these need to be accounted for. Companies need to look at the track record of a given vendor or channel – pay rates, cancellation rates and lifetime value. Unfortunately, most aren’t.
The lead generation firms have also been stealing share from the affiliate networks for years. Linkshare, Commission Junction and other affiliate networks have had trouble understanding their place in the value chain. For years they changed direction repeatedly, trying to be technology companies, then marketing companies, then search companies, then technology companies all over again. Meanwhile, they ignored their core offerings – tracking, payment, access to affiliates and overall accountability.
At the same time, other players, particularly the lead-gen companies, built their own technology. And they did so with lower overhead and focused only on the most profitable segments. The high-volume offers now appear on the CPA networks, and the top affiliates have moved to where the money is. Much of the remaining revenue streams have been under attack by Google AdSense and similar automated revenue-generation options for publishers.
The affiliate networks have responded by acting like CPA networks; Linkshare bought a lead-gen company in 2007 and is now trying to roll out its own lead-gen offering. But it might be too late. Tracking technology has been commoditized. You can even rent the technology and build your own affiliate or-lead gen network with just a few calls. I think we’re going to see continued consolidation in this space.
PERFORM: Given all the available tools for Web analytics, data integration and business intelligence, shouldn’t marketers be able to see which campaigns work and which don’t?
JACOBS: The tools and data are readily available, but marketers are simply not using them. I’ve seen multimillion-dollar online media campaigns run by top-tier Web properties with no tracking whatsoever. And I’ve seen millions of dollars’ worth of campaign-tracking technology go unused.
This might be a generational issue. Marketers coming from the off-line space not familiar with the metrics just don’t know what to look for. Many have outsourced the tracking function to IT, which makes as much sense as having coders develop promotions for your website. Campaign data does not belong with IT, it belongs with marketing.
Marketers need to take responsibility for learning the systems. More effort is put into setting up the tools and telling the boss they exist than learning how to analyze the data. People from the top-tier tracking solutions have told me that most of their clients don’t look at CPAs for their online campaigns. That is a very scary notion. Both technology and marketing dollars are being wasted.
PERFORM: As different service providers emerge to help companies with these various marketing initiatives, is there going to be consolidation, or are marketing departments going to continue juggling multiple third-party agencies? How do you see the competitors in this space shaking out?
JACOBS: Right now, I think the marketplace is too blurred. It is very easy to confuse who is who within the ecosystem. Companies of all sorts are trying to grab whatever share they can, regardless of whether or not they should. Conflicts of interest are almost the rule.
You have a lot of lead-gen companies whose pitches sound a lot like agency pitches. You have development shops treating campaign management as an extension of creative, when it really isn’t. You also have single-channel agencies trying to give broad strategic advice. Even the vendors and tracking solutions are getting into the mix, building their own account and campaign management offerings.
The worst trend comes from the vendor side. They don’t just want to sell you media, they want to “manage” your campaigns. This is a clear conflict of interest. Vendors are handling creative, talking about audiences and steering “strategy” in many cases. However, their goal is to generate high CPM [cost per thousand impressions] for their media, not ROI for their customers.
You also have agencies essentially outsourcing their interactive work to these vendors. Here you have the same conflicts of interest, yet clients don’t realize it because they think the agency is looking out for them. But the hands-off agencies rarely know enough to do so, or they don’t want to potentially risk their relationship with Google.
Other traditional agencies are partnering with channel-specific agencies, which leads to multiple teams managing different campaigns in silos. I’ve rarely seen this end in anything but disaster. With no accountability across channels, the campaigns often work against one another. The most frequently seen example is search campaigns conflicting with affiliates doing search. Done properly, with a view to overall performance, these are extremely complementary channels. Done poorly, you will drive up CPAs with no increase – and sometimes a decrease – in overall volume.
I think much of the challenge for traditional agencies is in adapting to a world in which everything is measurable and interactive. Online marketing – and performance marketing in particular – contradicts the approach that’s worked for those agencies for years. In the past, they just crafted one big message, segmented the market from the top down and then tried to cram the unified message through all the different pipes. This is push media.
The interactive space requires you to do a lot more listening than dictating. With pull media such as search, individual consumers can tell you exactly what they are looking for. Marketers can get unprecedented types of consumer feedback online, down to the campaign, minute or even specific search query. Companies now have data in droves, but rarely the experience and knowledge to use it.
In some ways, we’re moving backward. Some traditional agencies are falling back on older techniques that don’t actually work in the online space and setting the entire industry back. For instance, I’ve seen traditional agency types trying to overlay demographic information onto search without realizing they are making their campaigns less effective. Why guess that a site visitor is likely to be a golfer based on her age and income, when you can perfectly target ads to folks who are searching for golf equipment? The reason search costs more than other types of media is because it is almost perfectly targeted, so why revert to an inferior form of targeting and pay the higher media rates?
Too many people try to adapt what they know to the new paradigm, but some of it simply needs to be thrown out. Things are getting so bad that many of the media providers are making their products worse trying to adapt to these mistakes. MSN, for instance, now offers demographic targeting for search.
PERFORM: They’re basically throwing a useless tool at people who don’t know any better?
JACOBS: Yes. And there are other examples. I think day-parting – showing ads at only certain times – is another feature that has been sold in the wrong way. For instance, in a search campaign with a CPA goal, volume does not affect what a click is worth; conversion rates do. Yet MSN and Google are offering the day-parting option without the ability to see hourly conversion rates. Without the data, the tool is useless, but I’m sure it makes some traditional agency types feel better.
PERFORM: I suppose the positive spin is that this is one way of easing the old-school types into the Web 2.0 world, but it is a little misleading.
JACOBS: Right. It eases them in, but at the same time makes it tougher for those who are actually doing it correctly. This contradicts the goals of performance marketing and moves the conversation to the wrong place. Once a client hears a buzzword, it can be tough to get them thinking about anything else, even though the concept in question may not work for them.
PERFORM: Our whole conversation has been about the challenges facing online marketers, but can you boil it down to the biggest one for the next three-to-five years?
JACOBS: I think the fundamental question is: How do we integrate all of this data and all of these fragmented campaigns into one global brand? The different channels and media sources each have fairly specific attributes and customer feedback loops that we can optimize independently. But how do we match this up with trying to build and maintain unified brands? The days of marketing only being about full-color glossy pages and 30-second commercials are over. Now we might have just 30 characters to get our point across. It will be extremely tough to convey a full global message and graphic identity in a text ad on the right-hand margin of a blog. When the customer – not the advertiser – defines the marketing, brand takes on a new meaning.
There is no quick answer to this one. It’s going to take experience and data, and I’m sure we’ll see quite a few mistakes along the way. As I usually say, we’ve got to test, test and test again, and if that doesn’t work, test some more.