Strategy decisions through Share of Search
In our latest video with James Hankins, he explains why he thinks Share of Search is such a wonderful metric to create underlying data for strategic decisions, but how to go about that in practice? The most important is to understand what the impact was of particular decisions you and your competitors made. Here are some simple steps on how to obtain that overview:
Step 1: select the baseline
To maximise the insights from Share of Search, you need to compare it to your selected baseline. This is often the main metric you use within your industry. For the majority these will be linked to financial numbers, but it can also be other measures such as votes or infection numbers. The best is running regression analysis so you identify the impact of Share of Search, but a simple correlation showing the relatinship can also be a good start. The most common baseline metrics are:
Sales numbers: getting sales numbers on competitors can sometimes be tricky, but quarterly reports or official government databases can provide much information.
Market share: can be deducted from sales numbers, or often be obtained from industry organisations, such as the case for supermarkets in most countries
Opinion data: when having access to only certain time series of survey data, Share of Search can fill out the gaps in between
Step 2: plot the major activities
It requires a bit of industry knowledge to identify what major changes your competitors made. This can either be advertising campaigns or channels selection, but also changes in pricing, product launches or distribution channels. If you’re not certain about what to look for, there are often news announcements that can be of help. Should you not know when exactly they took place, the curves within the Share of Search can already give you an indication of what time period to look for. In most cases, it was some shift that caused people to actively start looking for something. IF you have access to search terms, that can also be of great help.
Step 3: correlate the impacts
Overlaying the major activities and your baseline, certain patterns will start to emerge. Regression analysis is obviously the standard to use here. Often it can be useful to work with different rolling averages to get a better view of the trends and not being too distracted by the different spikes. Check what drove the Share of Search to either increase or decrease? This can either be specific about one competitor, the entire category, or just simply show the big difference in impact between various competitors taking the same action. After all, every company is different, so a similar action can have totally opposite outcomes.
Step 4: build scenarios
Now that the patterns have emerged, you can start to build scenarios on possible moves your competitors have done in the past. What was successful for them, or which ones seem to be reoccurring? Start making different plays, based on their past behaviour. See what their responses could be and how that will impact the future Share of Search.
Step 5: bring it back to baseline and decide
Armed with the possible scenarios use the correlations and trends you’ve generated in step 3 to predict the impact on your business and evaluate what to do best. The longer your history, the more accurate your numbers will probably be, but that doesn’t mean you can predict further into the future. In general a 6-12month outlook is the maximum one can really do. Every type of “market”, whether it is business, charity or others, is after all also influenced by other factors such as marco-economics or social trends and not just by actions from you and your competitors.
If you use My Telescope, a big support in building these scenarios is the sentiment analysis with keywords and drivers. For those who don’t use it, Google news can be an additional support.