SKILLS FOR SUCCESS
January/February 2008
Measuring Online Marketing
Meaningful Metrics are Essential to Coping with the Speed of Change in Online Marketing
Donna Wray, Management Advisor, TGaS Advisors
With the pharmaceutical industry in an era of cost containment, brand managers face intense pressure to justify every dollar of their advertising and marketing budgets. Planning is essential to measure effectively and optimize results.
Analyzing the effectiveness of online marketing campaigns has come a long way. Marketing in the Internet Age has evolved beyond tallying total hits, unique and repeat visitors and the average length of time spent on a page. In the years since the pharmaceutical industry started treading warily into cyberspace, judging the return on e-marketing investment has become much more detailed and precise.
Today, brand managers should expect to receive a range of actionable information gleaned from their online marketing efforts just as they would demand from their traditional (offline) campaigns. The best vendors work closely with clients pre-launch to determine the most useful metrics for any given brand and deliver the information clearly.
The goal is to focus on the most meaningful information from the wealth of data generated. Collecting metrics doesn’t come cheap, making it all the more important to zero in on the measurements that drive the most program improvements. More should be invested in larger or pilot projects as they have greater potential to maximize the dollar value of learning and efficiency.
Advice on Choosing Metrics: Be Selective
Keep in mind two basic questions the metrics need to answer:
1. Did the campaign provide what was paid for?
The vendors need concrete accounting to prove they are delivering what they promised. This is the minimum metric.
2. Was the campaign the best possible for the money?
The most effective vendors are those that optimize over time based on what they learn in the market; they are able to sustain optimal performance even as some aspects become “tired.” They know automatically, through the metrics in place, what needs to be replaced. All this still doesn’t demonstrate whether a campaign is achieving the best possible outcome; for that, benchmark data are needed. Benchmark data based on similar pro-grams help gauge if an approach exceeds or is within an expected range, or if there is room for improvement by trying a new approach. Comparison information pro-vides a frame of reference for perspective and evaluation.
Metrics Choice Matrix
Choose metrics from the list below. For any metric chosen, select both volume statistics and efficiency (cost per “x”) statistics. Both volume and efficiency need to be taken into account because any program can be efficient at a very small volume, but inefficient as the program becomes larger. Benchmarking both ensures the proper balance and optimal market effect.
1. Rx Revenue.
Tying a program to actual incremental prescription volume is ideal, in theory. For a typical Rx product, however, it is not the best measure. It is usually handled via transactional data match. Because of HIPAA, this tends to be expensive and slow (six months or longer) and, since the data must remain anonymous, it rarely has much detail as to which parts of the program worked best or other information useful for improvement. Use this measure only if the program needs to be justified or com-pared to alternatives, such as a similar program offline.
2. Proxy for Rx Revenue.
Ideal, in most cases. The better a proxy, the more effective a tool it will be in optimizing programs. If a program has a rebate, coupon or voucher, then registrations for this can be a good indicator of how many real, on-target prospects have been attracted and convinced to start therapy. Since this sort of data is available to the brand manager, a much smaller sample (relative to measures through a transactional data match) is needed to compare different aspects within a campaign for faster, more effective optimization.
In general, a rebate is a better proxy than a coupon registration, but the regis-tration option yields faster results.
3. Second Choice: “Engagement” Stats.
Banner ads that tout “click here for a free…” might buy site visits, but they won’t necessarily get people who are good candidates for a treatment. The viewing of three or more site pages is the industry standard for measuring genuine interest; ensure each banner, sponsorship, etc. is linked through to determine which creative/venue/tactic choice yields the greatest number of interested visitors for the least cost.
4. Necessary, But Lower Priority: “Execution” Stats.
How much was spent, how many ads ran, clicks, etc. are all important to track, to ensure proper accounting and delivery against promises. All vendors must supply volume measures, but these are not useful to understand the value (ROI) of a program or helpful in optimizing the program.
5. Necessary Complement to Campaign Metrics: Benchmarks.
Benchmark data on what similar brands are finding successful can leverage existing industry experience, rather than learning after lengthy trial and error; ensure all useful opportunities are being explored; and ensure existing tactics are optimized against what is possible, rather than what’s been done so far.