Leveraging web analytics data is vital when it comes to optimally managing cost-per-click programs for e-commerce websites.
We're continually working to refine our own methods and systems, but have found that synthesizing CPC and Analytics data sets with offline accounting data has provided a foundation for solid management improvements.
With e-commerce accounts, we've found that there are a few important pieces of information that exist outside of the CPC interface, without which a program cannot be adequately (let alone optimally) serviced:
- Product margin
- Average order value
- Target profitability
- Customer loyalty / lifecycle
With product margin, the more granular the better. Although it can be extremely time consuming to provide margin information on a SKU by SKU basis for storefronts that carry thousands of items, this information is quite valuable. Many online retailers find it simpler and more manageable to provide margin information at the Brand (as opposed to the individual item) level; this is fine as well. At the end of the day, this information should be as accurate as possible. The data will probably never be 100% accurate, expecially given the propensity of many retailers to run promotions, but it should be reliable enough to build an effective bid management system around.
Average Order Value
Depending on the desired level of granularity in the management model, it is also possible to automate key management decisions by importing transaction-level metrics into the data set. Given that many orders will contain more than a single item, it is important to evaluate the total order when gauging the effectiveness of any single keyword.
By pairing transactions with keywords, and evaluating the average ticket associated with the transactions, and then extrapolating out the gross profit by subtracting the product cost, we wind up with a meaningful figure. From this point, we can continue and subtract the media spend associated with the transaction and have an even more meaningful figure.
When it comes to bid management, ensuring that the final figure is greater than zero is important. But a more thorough process will also evaluate desired level of profitability. If an organization is looking to achieve 20% net profitability, then the percentage of sale that goes to satisfy organization overhead, etc. needs to be accounted for as well.
Once the percentage of gross profit absorbed by overhead is determined, and the target profitability is established, a more complete landscape has been painted within which a CPC campaign can be effectively managed.
Customer Loyalty / Lifecycle
One final piece of the management puzzle combines (a) how many times and (b) how frequently customers return to purchase with (c) the client's desired business objectives for profitability. If the client wants to be profitable on the first transaction, well then customer loyalty doesn't matter. But if the client wants to be profitable - say - in the first 90 days of a customer acquisition, or by the second customer transaction, then these data points are very important components in the overall management strategy.
Apart from the difficulties associated with gathering all of the various pieces of information, it is frequent to encounter obstacles when it comes to marrying them together. There are a number of different ways that these obstacles can be overcome so as to automate the bulk of the management process, but - at some point - economies of scale will require that an API-powered management interface that integrates Analytics, AdWords and external data points is leveraged.
Couple this with muddled attribution models, and the whole process still boils down to a combination of art, science and judgment calls aimed at properly conjoining the two. But it's still a better model than most, and technology will continue to advance and unlock new possibilities.
Struggling to manage your CPC programs? Need assistance managing your analytics and bids? Let's talk - Salted Stone can help!