Do a search on a search engine or in your favorite social media sites and you will find many links to valuable advice on how to measure the impact of social media. Articles, columns, blog posts, and books all lay out the strategy and offer blow-by-blow detail of setting goals, defining key engagement metrics, optimizing analytics software packages, developing monitoring programs, building reputation management dashboards, crafting social media workflow solutions, and creating customized reports. There is a growing list of free or cost effective tools to measure social media, but how do you manage company expectations and report on results when what you are measuring is a first trial, you are working with a small budget, or don’t have analysts to lean on? Even free tools require expertise and time to learn that may not make sense when you are budget or resource constrained.
Real World Testing
Clearly, there is a chicken-and-egg problem inherent in tracking results in the test case scenario. In order for a test to yield actionable insights, you need to somehow gather those insights. At the end of the test you need to be able to determine whether the effort was beneficial to your business goals and should be funded and continued. But sometimes the scale of the test or the available budget rule out the resources needed to track, analyze, and report on the results. Sometimes you just can’t afford to do analytics full out, meeting all the best practices that we rattle off like it makes sense in every case. Sometimes it just doesn’t.
Real World Framework and Time Frames
Social media results often take time to build. Any good test needs to gather enough participation and cover enough calendar to allow for reliable trending data. It can’t be significantly influenced by seasonality, outside promotions, or other factors that would skew results, unless of course you are testing campaign response or other very specific situations. Expected results and the metrics you track will be heavily influenced by your starting point in social media, your goals, the type of program or campaign you envision, your market scale, your brand strength, and your intended audience demographics.
Once you are confident that the timeframe and other test specifics will yield readable results, then you can safely use your existing analytics frameworks and a little ingenuity to answer the most important questions about the impact of your test. At this point, you want to focus on the highest level indicators – no more or less information than you need to make a go or no go decision on the test.
Test Results Should Be Directional
If you finally get approval for a social media test but don’t want to dilute the impact by burdening the budget with lots of tools or analyst consulting fees, there are places you can skimp:
- Skimp on paid subscriptions to social media monitoring tools. While these are cost effective and beneficial for long-term programs, you don’t have that level of commitment yet. Start by ensuring that the goals of your program are well defined, then look in your current analytics program for relevant metrics. Set baselines before you start your test and measure on a regular basis during and after.
- Skimp on consulting fees. Develop a simple Excel spreadsheet to track metrics relevant to your goals using the simple analytics available in the admin sections of each platform. For instance, the “insights” tab on Facebook Fan Pages provides good information about who’s visiting your page, their geography, activity, etc. See if there is a shift in people visiting during specific campaigns; a shift might indicate an opportunity with a new audience segment. Take this information and cross check it with Google Analytics or another free analytics tool like Quantcast, which provides terrific audience information about who is visiting your site, including geography, demographics, “audience also likes,” and more.
- Skimp on daily or even weekly reporting unless the mechanics of your test direct otherwise. A beginning, middle, and end picture should tell you what you need to know without overburdening a simple test.
- Skimp on shades of grey. You don’t need to test multiple messaging points or creative versions. You don’t need to describe multiple levels of tonality to get a feel for audience response. Possibly the best measure of social media impact during a test is to poll your customer front line personnel like customer service or call centers to see if they notice any differences.
If it sounds simple, it should. I’m a huge believer in the power of effective analytics. Sophisticated programs require sophisticated data collection, analysis, and optimization, but I’m also comfortable with a few indicators when that makes sense and encourages smart testing. The best you are going to get from a short-term social media test is directional information. If the metrics are also directional, that is appropriate and enough information to allow for decision making. Of course, all interactive marketing, including social media, is an ongoing test, optimized regularly against the appropriate level of data. If this is truly a test, then treat it as such and match the right effort level to its measurement.
How many programs have you seen stalled or diminished by an overreaching analytics burden?