Search Data Analysis and Reporting: Providing a View of Global Campaign Success
Global search managers know that the lifeblood of a campaign is the data collected as the campaign is underway. As I mentioned in the first post of this series of points I presented at ad:tech in San Francisco, Peter Drucker has it right when he says “business has only two basic functions: marketing and innovation.” Well, innovation doesn’t blossom out of empty space in the SEO and search marketing world. There must be hard data that supports a new tactic or strategy. Managers must rely on analysis from dashboard reports covering a broad spectrum of data points to shape and direct search strategies.
When you look at the data that is captured through system analytics, the “5 Ws and 1 H” questions that form the basis of journalism should be posed.
- Why – Why analyze the data?
- Who – Who will be responsible for gathering data? Who will analyze the data? Who will the data be shared with?
- What – What are the risks related to analyzing data? What metrics matter most? What can team leads do to impact change?
- Where – Where will you find the metrics (which tools will you use)? Where are practices out of compliance with cloud protocols and/or search algorithms?
- When – When will strategies and tactics change? When (how often) will data be counted?
- How – How will governance be segmented? How will reporting be shared? How will response be managed?
The inherent challenge behind turning Big Data into actionable insight is two-fold: 1) You can’t manage what you can’t measure, and 2) all that is measurable should not necessarily be managed. So, you should measure everything but manage only the data that drives best practices in SEO and global search marketing.
There’s no shortage of SEO technical data that can be analyzed including content, site architecture, server delivery performance, and coding, to name a few. What’s important is to discern between correlation and causation when you analyze data. Data correlation reflects how closely two data sets are related. For example, your content scores well for relevance, so the page has to achieve high SERP ranking, right? No! Relevant content is necessary to compete for high visibility, but search robots include relevance as part of their 300+ ranking factors. Strong page relevance CAN be correlated with high SERP rankings, but that factor alone doesn’t cause the page to rank higher than other similar pages. Causation, on the other hand, implies one condition has a direct or indirect effect on another. Specifically, you can determine if you do X then Y will result—and you can then roll out that tactic repeatedly. Poorly translated title or meta description tags will directly influence lower regional search rankings and click through rates (CTR), for instance.
Each stakeholder has key metrics that affect the holistic success of SEO and search marketing practices. Following a campaign launch, each SEO team lead takes ownership of data sets within his or her purview. Web strategy, for example, is responsible for page rank, bounce rate, and formulaic metrics such as AOV and order rate; IT tracks crawl rate, 404 trends, and redirect chains; and global teams look at country Web rankings for key terms.
Teams can encounter data bias in their analysis process. Whether it’s relying on too small a sample size or seasonally affected data, stakeholders should guard against data bias.
Let’s talk about reporting. We like to work off an executive summary with a simple structure. Reporting should be focused on data trends. Year-over-year and quarter-to-quarter data show whether the strategies, tactics, and assets we’ve deployed have successfully moved the company forward and how we’re performing against forecast. Month-over-month data, on the other hand, reveal opportunities for agile managers to adjust practices swiftly. In SEO and search marketing, KPIs are about visibility and conversion. The executive summary should report visibility and conversion metrics with enough context to allow managers to understand factors behind changes in data sets.
In order to manage effectively, it’s important to report some data in multiple time frames. For example, we typically look at global revenue through MoM, QoQ, and YoY, which drives sharper forecasting. And don’t forget this: data sources should be noted in the report. Remember causation? Providing data sources in your reporting process can help pinpoint causalities on the way to adapting practices to meet market reactions.
At Adobe, we look at KPIs like visits, subscriptions, revenue, trial downloads, and orders on a monthly basis in order to respond quickly to successes and failures. Often, we analyze correlative metrics such as visits to revenue ratios (RPV) that, while simply providing a 30,000-foot view, indicate whether analytics, e-commerce, Web strategy, or other SEO team members need to take a deeper look at the data.
I’ll leave you with this thought: Big Data isn’t about having more data to sift, sort, and distribute. It’s about reaching for the data that matters, identifying patterns that support or refute whether your practices are successful and anomalies that may point to potential issues, and then executing the strategies and tactics that support global search marketing success.
In my next post, we’ll wrap up our six-part series with a discussion about scalable planning.