Attribution and Brand Awareness: Determining the Effects of Your Marketing Activities

We all know brand awareness matters a lot, but it is always difficult to attribute conversion value to our various marketing activities.

So how can brand awareness activities be accounted for? Online attribution tools do a great job of measuring online activities, but what about attribution with offline activities?  

To do this, marketers need to identify all regular marketing activities and then establish a baseline.  The baseline could be established using Google Insights for Search (as described above), if the brand generates sufficient search volume to show up.  Another way to do this might be through a lead tracking system as long as all leads are tracked in a single system that is readily queried.  While it’s probably best to track actual sales in determining a baseline, for some businesses the sales cycle is long, so establishing a baseline by using lead volume may be just as good a proxy and provide a more immediate data set.  Whatever the method, adjustments for seasonality should also be made if appropriate.  

Once a baseline is established, then the marketer should ensure that all activities (except those being tested) remain constant.  This includes minor changes like changing your target audience or adding a new media type. With a baseline established, marketing activities held constant, and a fully integrated attribution system in place, test campaigns can now be run to provide marketing insights.  When multiple media types are to be combined into a single campaign, the attribution system should understand how to allocate conversions to the media types both independently and together.

In the example graphed to the right, we illustrate a simple example of how a marketer might properly attribute conversions when adding a paid search campaign after establishing a baseline using lead volumes.  

Even though a search engine marketing campaign generated 150 out of 300 leads (an apparent 50% of the lead volume) in the test month versus a total of 200 leads in the baseline period, search engine marketing should only be attributed 100 leads (net increase in leads from baseline) or 33% of the value of the conversions in the test months.  

How could this be fair?  After all, the search campaign generated half the leads in the test period.  First, this is fair because month after month the business always generated approximately 200 leads (adjusted for seasonality).   Second, the search campaign may have captured the customer before she used some other source for attribution.  Because the prospect already was aware of the brand, had she not found the brand via search, she would otherwise have used another source to find the brand.  For example, a prospect could find a brand via a free, local business listing, as opposed to paid search.

If new marketing campaigns are to be continually tested, marketers need to clearly identify what baseline is before proceeding, noting that there may now be a “new” baseline.  If the search campaign above is terminated and replaced with something different, then the new baseline would remain the same.  If, however, a new campaign is layered upon it, then the new baseline would be 300.

As the number of marketing tools and channels increase, the more difficult it will be to attribute revenue and brand awareness to our activities.  How do you track your marketing activities?  Have you come up with any solutions to the problem of attribution?

Supporting Multiple Languages with Marketing Automation

Since many of our clients are global brands, our local marketing automation platforms must support multiple languages. When supporting multiple languages, the more you know, the more you realize the difficulty of the task at hand.

One major challenge we initially faced: varying lengths of copy depending on the language in which it appears. A paragraph in English tends to be much longer when written in German or Japanese (see the image below), and much shorter when composed in Simplified Chinese. So how can a marketing automation system account for that problem?  We have had success using line based copy-fitting rules, but we think this is an area where the marketing automation industry, as a whole, can improve.

A second challenge we face involves a brand's licensed font set.  Most major brands choose a font or set of fonts that their affiliates must use for brand compliance.  Clients  often choose their branded fonts for western characters, but those same fonts might not be available for character based languages.  This is a discussion we often have with our clients who may not initially realize just how many font licenses they will need to purchase to ensure global brand continuity.

A third issue to consider here is the accuracy of translations between languages and the difference in social connotations associated with words or phrases.  We've all heard stories about phrases that are funny in English but translate into something meaningless -or worse, offensive- in another language.  What is the best way to build and maintain a marketing automation system that regulates the literal and social translations between languages? One way Pica9 addresses this issue is through our approval loop feature, which you can learn more about here.

The last major challenge we see in maintaining multi-language support for local marketing automation systems is communication amongst participants.  We feel it is up to us to help foster communication between our clients, their local affiliates, and our in-house team of developers and customer service representatives.  We approach these situations with a great deal of patience, but still, this is perhaps the most challenging issue of all.

What are your thoughts on carrying a brand across multiple languages?  What are some other major challenges that we have not mentioned here?  Do you think this is becoming less of an issue as we experience globalization, culturally and economically?

Digital Asset Management Systems: Cost vs. Customization

Digital asset management (DAM) systems come in all shapes and sizes. They range drastically in price and also in their customization capabilities. Usually, the more customizable the system, the heftier the price tag.

As we can see from this graph, the level of customization is directly correlated to the price of the system, with few outliers. However, in our experience, this relationship represents an inefficiency in the current DAM system market.

A low cost, out-of-the-box system can be problematic because the creators of the system make broad stroke decisions that greatly effect the usability of the system.  For example, the system predetermines types of users and their respective level of authorities in an out-of-the-box system.  But is a predetermined organization chart realistic when no two companies are exactly the same?

On the other hand, we have noticed several problems with completely customized DAM systems, built from scratch, for specific customers.  As you can imagine, creating a DAM system is a complicated process that requires a great deal of time and work to ensure it runs smoothly and bug free (hence the heavy price tag).  So companies who purchase custom systems are paying to reinvent the basic platform, but the end result is a less stable system that takes a long time to develop.

The solution to these dilemas is far from simple, and we continuously explore and discuss the possibilites.

As a result of this exploration, we have been working on a better solution that breaks the linear relationship between cost and customization.  As seen in the picture to the right, we've built, tested, and troubleshot 80% of the our system already, and this becomes the base to every DAM system we implement. We leave the last 20% for when new clients sign on and can give us the information we need to customize the system to their needs.  The result is an affordable, mature, custom DAM system for our clients.

What are your experiences with DAM systems?  What features are most important to you?  Do you prefer an out of the box solution, or a customized solution?  Do you see a market for something that speaks to both ends of the spectrum?