Attribution management is the practice of allocating credit (revenue, profit, etc.) across the team of advertising responsible for a conversion (sale, lead, download, etc.) versus giving the very last ad clicked all of the credit. Any good Internet marketing professional knows that assigning all of the value to the last ad clicked before the conversion is inherently flawed. Most Internet advertisers, however, lack:

1. The tracking technology required to determine the actual team of ads and their sequence that lead to the conversion, and

2. The valuation methodology to properly assess each ad’s true contribution and value to the conversion.

There are a few technologies emerging that are able to effectively track and assemble the team of ads leading to conversion, including our own Purchase Path^TM. technology. With over 2 years of market presence with our attribution management technology, we have been able to capture millions of “purchase paths” across dozens of companies and numerous industries. The remainder of this blogs highlights some of the major “learning’s” and insights we have been able to draw from the massive amount of attribution management data we have collected.

• Branded Keywords:
  • Whether you are an established or new brand, an eTailer or lead generator, in the B2C space or B2B space, the most common Purchase Path we have observed is where the last ad prior to a conversion is a branded keyword from a search engine.
  • Ask yourself, what is your best performing keyword? I would be willing to bet that your answer is some derivative of your company’s name. The reason why companies believe these branded terms are best is due to their inability to track and view a Purchase Path. Our data has conclusively proven that users are generally not searching for your brand at the start of the buying cycle; rather they are searching for the types of services and products you offer. It is during these searches that they discover or become acquainted or re-acquainted with your brand. Once they get to this point, the user may look at other options, or your competition, and after doing so, they “navigate” back to the site that has the best offering and value. The most common way to “navigate” back to a site is to type the company name into a search box and click on a paid search advertisement. It is this common search behavior that inaccurately inflates branded keywords and makes companies improperly conclude they are their best form of advertising.
  • Because of this branded keyword bias that exists today, your other forms of advertising are often undervalued, such as banners, emails, and category-level keywords that generally occur early in the customer buying cycle.

• Window of Time:
  • Shortly after we developed our Purchase Path technology, the question of ‘How far back in time should we track the path when doing attribution management?’ arose. For example, if you have a website that sells toasters and someone clicked on an ad 2 years ago, then clicked on an ad today and bought a toaster, technically that is a Purchase Path. However, most marketers would agree that the ad clicked two years ago had no impact on the sale that occurred today. Given this intuitive issue of time relevance of an ad’s effectiveness and impact, it becomes imperative that any advertising analytics technology that focuses on Attribution Management needs to include a time-based parameter that controls how far back in time one should go when assembling their purchase paths.
  • One method that we developed to help answer this question analytically is to plot your sales on a spreadsheet from the time of first click to sale. Once a large enough sample of your sales has been captured to be statistically sound, take a look at the time that it took to get 80% of your orders from first click to buy and also look at the time it took to get 90% of your orders. You can then rationally set your Purchase Path time parameter as the average time it took to get between 80% and 90% of your orders.
  • We have found that one time window does not fit our entire client base. The types of products/services you sell, their price points and average sales cycles are the three biggest factors that will influence the appropriate purchase path time parameter for your company.

• Introducers, Influencers, Closers:
  • Our Purchase Path tracking technology showed that it often takes more than one ad to earn a customer. For most of our clients, over 50% of their customers clicked and/or saw more than one advertisement before becoming a customer. To help understand the interactions and roles of ads in a Purchase Path, we created some base classifications to categorize these ads:

§ Introducer- the very first ad the prospect saw or clicked

§ Closer- the last ad clicked prior to the conversion

§ Influencer- any ad that is seen or clicked between the Introducer and Closer

o When we work with a new client, we find that the majority of their ads fall in the Closer ads, and they have very few Introducers and Influencers. If you think about it, this makes a lot of sense. Companies can only justify buying advertising that demonstrates it has value. Without the benefit of being able to see a Purchase Path, the only ads that appear to have value are the Closing ads. The introduction of this classification system allows us to rationally group ads into logical buckets and in doing so, more accurately value them, no matter where they appear in the buying process.

• Comparison Shopping Engines and Search:
  • Perhaps the form of advertising that is undervalued more than any other is Comparison Shopping Engines (CSEs). CSEs more often than not fall in the Introducer or Influencer stage. We have discovered that consumers use CSEs to compare prices, click on the listing they deem to be the best, leave the site that had the best offer, then go to a search engine and search for that same company by name (branded term). From there, they click on the company’s branded ad and complete their purchase, thus giving all the credit to the search engine and none to the CSE. Our analytics do not tell us exactly why consumers do this, but one theory is that consumers are uncertain as to how CSEs make money. Consumers may believe that CSEs jack the price up a bit; therefore, consumers think if they go directly to the site through a branded term or direct visit, they may find a lower price. When they realize the price is the same, they do not go back to the CSE and simply remain at the site to make their purchase.

If you are a marketer that is currently tracking the Purchase Path and performing attribution management, then we would love for you to share your findings to our own data. In effort to advance the knowledge of attribution management, ClearSaleing will be co-hosting a webcast with Search Marketing Now on October 28th that will survey experienced online marketers about different attribution models, or various ways to properly allocate revenue and profit to each contributing ad in a conversion. If you would like to understand how other marketers are handling attribution, please sign up for this webinar: www.clearsaleing.com/attribution.

If you are not using some form of Attribution Management when evaluating the performance of all the elements of your marketing mix, then you have an even greater opportunity of maximizing the ROI of your marketing investment. If you are applying attribution management principals to your performance evaluation, you have made the move to the top of the online marketing sophistication food chain.

While your competitors are still attributing conversion credit to the last ad clicked, you are attributing credit to the entire team of ads that led to the conversion. This new measurement methodology is going to allow you to:

• Conclusively prove the value of certain types of ads, like banners and emails, which typically occur earlier in the buying cycle.
• Properly attribute value to your search ads that, up until this point, have probably received too much credit.
• Make smarter media buys and invest in new forms of marketing that previously could not be justified.
• And most importantly, to increase the profits earned from your online marketing budget.

Seems simple, right? Unfortunately, like with any new endeavor, the devil is in the details, and Attribution Management is no different. If you are unaware of some of the key factors that influence the Purchase Path and your ability to perform Attribution Management, your new initiative may not be as impactful or successful as originally hoped. There are 3 principal pitfalls that, if properly addressed, will go a long way to ensure that your Attribution Models will be more accurate and that you will be much more likely to enjoy the benefits that Attribution Management can deliver.

Double and triple counting conversions or worse:

One of the biggest risks with Attribution Management today is the potential to count more conversions than actually took place. This can happen in several ways, including:

• If you have more than one conversion tracking code on your site. For example, if you have Yahoo and Google conversion tracking, and a person clicks on a Yahoo ad, then clicks on one of your Google ads and then converts, both engines take credit for that. The Yahoo conversion tracking does not know that Google was also involved and vice-versa.

• Larger companies typically have a person or team dedicated to search, another team for display, and another for SEO, etc. Each has to generate reports to show the value of the work they do. If the display team can show that a particular conversion that ended with search also was shown a banner, they will want credit, but so will the search team. If, in that same conversion, an organic search was performed, then the SEO team is going to want to take credit, as well. When you have multiple teams all trying to justify the work they do, you are likely to end up with inflated conversion counts.

Solution: Instead of using the tracking tools provided by the ad sources, invest in or develop your own tracking technology. The use of a 3^rd party or in-house advertising analytics technology will prevent the double counting and will prevent various teams in your organization from taking individual credit for the same conversions. To go one step further, you should standardize the metric you are using to measure and attribute performance across the team of ads, which leads to the next point…

Not attributing profit:

Another big problem in the application of Attribution Management at many companies is what and how they actually attribute and assign value across the contributing team of ads. You need to ask yourself the following questions: What is a conversion worth? Are 10 conversions better than 1? Are some conversions better than others? Are all conversions profitable?

• A conversion is valuable to an organization if it produces profit. A conversion that loses money for a company is not desirable (though one could make an argument for loss leaders or Life Time Value (LTV) here). In order to truly determine if a conversion is valuable, you need to dig for more information. Specifically, you need to know: (1) What did it convert? (2) How many advertising dollars were needed to generate that conversion? and (3) What was the margin generated from the conversion?

• If you are able to answer these questions, then you can start attributing profit vs. conversions to your team of ads. Or, if you want, you can do both. Profit does not lie. If an ad is profitable, it is worth keeping and investing more money towards. If you only know that an ad produced a conversion, you still do not know if it is profitable and therefore, worth keeping.

Solution: When deciding on an advertising analytics technology that allows you to perform Attribution Management, make sure it is able to track more than just the conversion or revenue generated from the conversion. It must be able to track profit. And in order to calculate profit, the advertising analytics platform needs to know which products were purchased as a result of the conversion, their respective margins, and the cost of all the advertising that led to the conversion. Your marketing meetings will be greatly improved when you can definitively say you spent X dollars on advertising, which generated Y profit for your company, versus saying you spent X dollars on advertising and generated Y conversions. You have probably already been asked the questions that more and more CMOs and CFOs are asking their marketing teams: “What did we spend? What did we convert? How much profit did that produce? And what is our ROI on that expenditure?”

Time

The final of our 3 pitfalls relates to the time period over which companies actually assign attribution. You need to consider several questions when assessing the proper timeframe for attribution management: How far back in time should I look when doing attribution? Should I credit ad clicks that may have occurred a year or two ago if I can track them, or should I focus on a shorter window of time? How can I figure out what the right window of time is for my company?

• Your answers to the above questions depend largely on the type of business you are in and the associated length of your sales cycles.

• The attribution management tools provided by the search engines are very limited when it comes to customizing the length of the attribution cycle to match your specific business attributes. The engines build one-size fits all tools and the one-size they fit is what fits best for the search engines to make the most profit.

• Some of the engines show the number of