One of the first things listed on any PR measurement post is the need to establish benchmarks and set goals for measurement, if it’s to be effective. Knowing you need to set these things in place before proceeding with a measurement program is one thing, setting out to do it is another.
Here are a few tips to get you started on identifying what needs to be benchmarked, and how to set goals that will be measureable and attainable.
What can you measure?
- Sentiment and tone – Establishing a benchmark can help you see how attitudes toward your brand or company change over time.
- Volume and number of mentions – Tracking volume is particularly helpful if you are launching a new brand, product, or other new-to-market item. Overall generic volume is less helpful to measure. One exception to this is in a crisis—in that instance it is very useful to see when crisis mentions drop either due to PR efforts, or when the crisis is over.
- Topic mentions – This could also be called “message mentions” or “product differentiation mentions.” Tracking specific messaging in conjunction with a brand name helps to demonstrate how and where your brand or product’s key attribute mentions are taking hold.
- Share of voice – I’m a big believer that SOV should only be tracked in conjunction with sentiment or tone. As mentioned in a previous article, BP might have garnered an enormous amount of the SOV of oil companies during the Deepwater Horizon disaster, but I doubt anyone would approach a client suggesting a big PR blitz to match that SOV. Devoid of some kind of context, SOV can lead you in some wrong directions in PR.
Setting up a baseline
For any of the items above, you’ll need to pick an amount of time to view for the baseline. If you are already tracking these items, identifying your baseline should be fairly straightforward.
- If you’ve already been tracking for a year, six months, or a quarter for sentiment, you’ll want to identify any segments of time that should be excluded from the baseline.
- This means, for example, removing periods of overly positive coverage and overly negative coverage. If your client made a large donation to a hospital and had weeks of positive coverage, you’ll want to exclude that from your calculations.
- Do the same with any horrible coverage, like the BP example above.
- Why? Because these bouts of intense ups and downs will make your baseline less accurate. If you leave in all of that positive coverage, your baseline will be artificially high and setting goals based on that (say, to improve positive mentions by 10 percent by the end of first quarter) will be inflated and hard to reach. Conversely, leaving a slew of negative mentions in will make it easier to reach an “improvement” goal, but it won’t reflect reality.
- Examine what remains after removing extremes, and find the average percentage of sentiment for positive, neutral, and negative mentions. That is your baseline.
- Repeat the process for volume, mentions, and SOV/with tone.
If you are not already tracking these items, you’ll need to collect data in order to establish a baseline.
- Determine how much time you are willing to dedicate to monitoring simply to establish a baseline (a week, a month, or a quarter);
- Gather the data;
- And then perform the analysis above. Remember that the more data you have, the more accurate your baseline will be.
Once you have your baselines established, you can move forward on setting goals for improvement that correctly map to your measurement program.