Cross Pixel

The following was originally posted on adexchanger.com:

I have been overwhelmed by Zappos retargeting ads from Criteo recently.  Apparently, I am not the only one.  Michael Learmonth wrote a piece in Ad Age yesterday about the subject and I think it about it a lot (see post).  Data driven marketers are walking a very fine line, balancing what we can do and what we should do, and I believe that Criteo and others that focus on automated, dynamically personalized  retargeting creative  have crossed it.  What’s worse is that there is no need to cross this line because I don’t believe that this approach to retargeting improves performance and I know it pisses a lot of people off and can hurt a brand.

I am a big proponent of retargeting and we do a lot of it at Cross Pixel.  It is an effective strategy.  My concern is with retargeters that use tactics that scare the consumer by exposing too much information in the ads.  One popular tactic is to utilize the last products searched as the basis for the ad creative.  The theory is that the ad will perform better because it is being personalized (dynamically) to the specific products that were of interest.  It makes sense, in theory, but it can be annoying to consumers in practice.  Since buying a new pair of very cool Puma sneaks from Zappos about 2 weeks ago (the author is going retro), I believe I have seen 300+ ads from Criteo for Zappos that continually show me the same 5 pairs of sneakers that I was researching.  Here are some suggestions that retargeters should consider:

1. I already bought the sneakers they keep pushing, so I would recommend that remarketers remove buyer’s cookies from these types of retargeting efforts.  A consumer that isn’t in the business would think that Zappos is a clueless company (“Why does Zappos keep bombarding me with ads for a product that I already bought?”)
2. Frequency caps are even more important when you are dealing in personalized ads.  I have seen the Zappos ads 10 times or more on certain days.  Most marketers utilize a frequency cap of 1 or 2 imps per day per unique.  Dynamic retargeters MUST utilize frequency caps.  My sneaker purchase was one small part of my life about two weeks ago.  Based on the frequency I see the ads from Zappos, it feels like that one shoe buying event is defining my online existence.
3. Dynamic, personalized creative can produce as many false positives as positives.  If I was on an auto site and I looked up lots of Ferraris and Lamborghini’s does that mean I am “in market” for those cars?  I wish that were true, but it isn’t the case.  Dynamic creative can make many mistakes – the last products you looked at aren’t necessarily the ones you will react to and buy.  What if the last products you looked at were a total turn off and made you leave the site? This is a risk not worth taking.
4. Overkill – this is my biggest concern.  Based on experience, automated, personalized creative in retargeting efforts does not improve performance.  The reason this is the case is because you are targeting a consumer that is already predisposed to the web site that they visited.  It doesn’t take much to get them back to buy goods at the site.  Simple ads that remind them to return are all you need to get people to return and buy.  I equate it to Search Engine Marketing and bidding on your brand name.  When someone goes to Google and searches for “Zappos, they are likely going to click on the Zappos search result regardless of what the text ad says – they were already predisposed to Zappos.  There is really no need for an SEM firm to focus on the text ads for a “Zappos” keyword search.  They don’t need to personalize them, or test varieties of copy.  There is only so much fine tuning that is necessary – optimizing the ads for the keyword Zappos is a total waste of time and the same is true for retargeting. It is highly likely that the consumer will return to the site they already visited if they are interested.  Personalized creative is overkill and unnecessary.

I am sure the retargeters using these tactics will disagree with me and I look forward to hearing why I might be wrong.  Our data says automated, personalized creative doesn’t improve performance in retargeting and I know I am not the only one who is annoyed by this type of advertising.  The data driven marketing community needs to show a little restraint if we are going to win over acceptance by the consumer.

Can a single DSP give an advertiser access to all of the desired impressions available on the ad exchanges?  I used to think the answer was yes.  I thought each DSP had access to the same inventory and that each DSP had the capability to process ALL of the available opportunities from the exchanges.  Ahhh, the naïve days of the early DSP era (my thinking was so 2009).  I posed this question to most of the DSP owners that I work with and I am pleased to report that all of them were honest about the fact that their DSP couldn’t make every impression available to a single advertiser.  The reality is that a single DSP can’t connect/ interact with the exchanges in a way that gives an advertiser full access to the inventory they are seeking.

So what is an advertiser to do?  Use multiple DSP’s? (Yes). And what does this mean for potential DSP owners?  If you are large holding company thinking about buying a DSP, doesn’t it reduce its value to you if you can’t build your entire “trading desk” around one DSP?  I think the answer is yes.  It seems to me that agencies need to focus on building a demand platform that has ability to integrate with the entire pool of available inventory.  This will require working with multiple DSP’s, multiple ad networks and directly with publishers.  For a demand side platform to be really valuable, and a single solution that can meet an advertiser’s needs, it needs to have access to all of the inventory in the marketplace (or as much as possible).

Imagine if a Search Marketing platform didn’t allow you to bid on all of the keywords that you wanted to purchase.  You would end up using multiple platforms.  Unfortunately, this is where we are with DSP’s.  Will the answer be that DSP’s develop the capability to access all the exchange inventory, or will they end up differentiating themselves in another way? My guess is the latter.  It is probably time to say goodbye to the concept of totally neutral platforms.

My birthday was last week and I got the usual assortment of phone calls, texts and Facebook shout outs from family and friends.  I must admit that I don’t put the same value on a Facebook “Happy Birthday” wish as I do from someone that calls me or texts me.  It’s not that I don’t appreciate a FB note but if all you have to do is logon to FB, notice its my birthday and type a note in 2 seconds, it is a wish that is devoid of some of the qualities that make a personal birthday wish special – that someone goes out of their way to remember your day and makes an effort to let you know.  FB has commoditized “Happy Birthday” and changed its meaning to me.  Not bad, just different.

I think the same can be said about the FB “Like” button.  “Like” has made it so easy and effortless to tell everyone what you “Like”, that it diminishes the value of the intent.  Effort is a key component in determining the level of a person’s interest.  A lead form that asks for detailed information for a life insurance quote will yield much higher quality leads than a life insurance form that asks for three basic pieces of information.  Effort weeds out lower quality and lower interest.  As “Like” becomes ubiquitous, it’s value as an intent indicator is lower.

This is what I think this means for FB in terms of monetizing “Like” – it won’t be the new adwords (the home run they are hoping for) but it will be a leading source of data for audience targeting (still a decent opportunity).  If FB really wants to turn their data into a powerful solution, they need to harvest real search intent, either on site, or off site – which I think is still very possible and likely to happen.

Do you remember the first few years in online media buying when the major print publishers offered banner ads to their advertisers as a value ad?  It wasn’t that long ago and it didn’t do anything to help save the newspaper industry.  While most of the talk is about the decline in subscribers and readership, I believe that this problem is secondary to the newspapers key problem – their inventory was overvalued by a magnitude and they could get away with it because they controlled distribution and had “exclusive” access into our homes.  We all know those days are gone, but there is still a fairly robust group of people that like a physical newspaper and haven’t shifted to reading the local news online.

Now the newspaper publishers that are successful online are facing a different issue – they have way more inventory than they can sell directly to advertisers at full price.  So what is an offline/online publisher to do to compete with hundreds of other news sites and brands that capture our online attention?  I suggest they start throwing in print as a value ad.  Print advertising is still very valuable, and I believe that the newspaper industry could use this as a key differentiator to garner a larger share of full price, online display advertising.  I know that my clients would value newsprint ads if they were included in a package and I am sure it would lead them to more buys from newspaper sites.

I think this is inevitable, but it probably won’t happen on a major scale for a while.  When print becomes a value ad it will be the final confirmation of the end of newspapers as a print advertising medium and few newspaper owners are willing to accept this fate, even if they know it is coming.   I suggest they learn from the past and jump ahead of the issue and use it to their benefit.  I would be interested in learning if any large advertisers have received this type of an offer from the newspaper companies.

I believe that right pricing model for an advertiser to execute retargeting campaigns and other targeted display campaigns is usually CPM.  While it is intuitive to select CPA as a pricing model because it ensures a specific performance, the reality is that buying on a CPA often limits the scale of your campaign. CPA buying shifts the risk and the responsibility for the arbitrage from the advertiser to the media seller.  The typical reaction of a media seller in this scenario is to limit it’s risk and only pursue media buys that are performing well below the CPA earn out.  For example,  an advertiser is willing to pay $50 for every sale generated during a retargeting campaign.  The retargeting company starts buying media to determine what the right price is for the inventory.  Because the retargeter is taking all of the risk, they will typically set their allowable ad spend, i.e. what they are willing to spend on advertising per order, at a low level to make sure that they do not lose money.  For this example, the retargeting company might set the allowable at $25 so they can try and earn a healthy markup on each order.   The problem with this model is that the retargeting company has very little incentive to push the envelope anywhere near the real allowable.  The basically go after the low hanging fruit, make a very high margin, and avoid the risky arbitrage.  In this example, the company may have figured out that a $1.50 CPM backs into the $25 allowable for the retargeting company.  The company will not bid on any inventory above $1.50 and the advertiser suffers because there are many orders that can be acquired by bidding above $1.50 that will fall within their $50 allowable.

CPM buying is much more scalable because it eliminates the arbitrage.  The retargeting company in a CPM scenario is working on a smaller fee and their portion of the revenue is tied to the media spend.  In a CPM scenario, the company is motivated to help the advertiser  maximize the number of sales they can generate within the allowable.  In the example above, a CPM retargeter will bid well above $1.50 to make sure the client gets all of the orders possible.  In a CPA scenario, the retargeting company is motivated to maximize its own profit margins.  This fundamental difference results in many lost sales for the advertiser.

Many advertisers are still concerned that the CPM based programs motivate media companies to spend, simply because that’s how they make money, while they don’t have to worry about this in a CPA buy.  White it is true that the CPA buy has a built in regulator, CPM buying can be easily controlled through daily reporting and the reality that every smart performance oriented company knows they have to meet the CPA goal to satisfy the client and keep the business.  As long as you know the company and trust them, an advertiser is almost always better off buying on a CPM because it can lead to many more sales , particularly on high performing programs like retargeting.

The two most valuable data points in online marketing, in terms of delivering ROI for a client, do not require a complex alogorithm or a secret sauce.  When someone conducts a search, they are giving up one data point.  Based on that one data point, a buyer can place a related ad and generally get a good return.  In the Audience Targeting space, the same is true.  A site visitor provides one data point - a browser reached a product page on a site.  Based on that one data point, marketers are generating terrific ROI executing retargeting campaigns. My point is that the two most effective online marketing tactics are based on a single data point.  No doubt there are algorithms that can measure the effectiveness of these single data point activities, extrapolate and increase performance, but the primary efficacy of these strategies centers around a single data point.

Of course this doesn’t mean that a sophisticated algorithm based on the analysis of many data points can’t work, but in most cases advertisers don’t buy enough media to provide enough data to make the algorithm reliable and actionable.  When you are analyzing volumes of data points, you need a lot of volume to deliver actionable information (I’m sure that comment will not sit well with some of the company’s developing predictive models based on limited data, but I have never seen this strategy work for a client and I remain very wary).  There was good post I read this morning on Cogblog that illustrated my point – even at Ad.com, arguably the biggest ad network out there, with volumes of data, more data is needed to improve the algorithms.  In his post, Brent Haliburton argues: “I worked at Ad.com and I am not gonna lie, I took away from that place a sense that algorithms are hard, you need tons of data to figure out how to improve them, and there are few shortcuts other than “been there, done that”. Unlikely that any small company has a better algorithm today.”  I’m no techie but its hard to fight Brent’s logic.

I have been focusing on this subject because I fear that algorithms have become a smokescreen for fraudulent operators and advertisers need to be wary.  My advice to ad buyers and online marketing people is this – before you place an order with a new DSP or Ad Net that relies on their secret sauce to deliver results – just make sure that the secret sauce isn’t View Conversion Spam – because it usually is.  When you have had a lot of success with single data points, it’s easy to be cynical.

More data, another DSP and another snappy startup that has some kick ass algorithm that will change the world for marketers.  Just another week in the new display ecosystem that is emerging around the seperation of media and audience.  I’m sitting in the middle of this thing, and I can’t help but wonder if the people that are creating all of these new businesses are doing the most inportant research that they can do – talking to potential customers to see what they actually want and need.  It seems obvious – create products and services that meet a need – but all too often entreprenuers don’t think that way (sometimes for very good reasons).  It’s the only way that I think and I am hearing things from clients that some of the new service providers would benefit from hearing too – algorithms are bullshit and they are often a red flag for media buyers.  OK, of course some algorithms are valuable and effective, but I would argue that most vendors that have  a product that relies  on a “proprietary algorithm”  have nothing at all to offer and they use the algorithm to make their offering sound compelling.  Well, I have some news from the field – buyers aren’t buying it.  They want transparency and they want understanding.  If your pitch involves asking the media buyer to trust you, you are not going to win many deals.  If you can’t explain exactly what your company does and HOW they do it, you probably won’t win many deals.  I think the winners in the new data/media space will be the companies that develop the managed services and products that are understandable to buyers.

In my opinion, there is too much focus on infrastructure and not enough on service.  If the whole data driven display market turns into a $3 billion category, how much money can be made from infrastructure?  20% of the total market?   The money will be in the managed services and fancy algorithms will only go so far.  The media buyers are looking for real products, not magic.  I plan on meeting their needs, not dazzling them with bullshit.

I am about to hire many sales people and I have never hired a large sales force before.  It started me thinking about the type of person that I would want as a sales person.  I remember a conversation or two early in my career when people told me about a sales person that could “sell ice to an Eskimo”.  They said it with pride.  I am a pretty good sales person and I know for a fact that I cannot sell ice to an Eskimo.  They don’t need ice.  I wouldn’t even be able to start a conversation with an Eskimo about my fine line of ice products.  I would be embarrassed to broach the subject.  To be good at selling ice to an Eskimo, it seems to me, you have to be good at lying.

On the flip side, I can sell something I believe in really, really well.   When I stand in front of a client and I know I am offerring them something that they need, my confidence shows through and I usually make the sale.

My conclusion – I am going to search for honest people that believe in my product.  And, of course, are willing to work their asses off to make a lot of money.

So…if you understand the impact of the separation of audience and media, and you understand why previous shopping activity is the leading indicator of future purchases, and you can’t sell ice to an Eskimo – please contact us, we may have a great sales job for you.