Share of search (SoS) is the fancy new kid on the block in PR measurement. It’s travelled over from the world of SEO, and is being trialed by PR practitioners to measure how earned media is impacting online search behavior.

SoS measures the volume of internet searches over a specific time period for a branded keyword, compared against competitors. On a granular level, this gives PR practitioners the ability to measure coverage impact by measuring how many people use search to research a brand following a major announcement.

To my mind, SoS is a massive leap forward in PR measurement. It means that we can more accurately see when we’re generating awareness for an organisation and measure how it’s improving over time. It’s a golden goose metric that measures impact, and proves PR’s value without muddying the waters with vanity metrics, internal attribution metrics, or AVE.

As always, there’s a catch. It doesn’t always work, and if used incorrectly can cast earned media in poor light.

SoS can be used to measure brand awareness without having to spend thousands on surveys, but it’s not perfect. There’s no real way to measure how many people are searching for specific keywords, though there are good tools with good estimates. In some ways it’s similar to using Google Trends but with more granular insights into search volumes, making it much more valuable — but still to be taken with a pinch of salt.

It’s also not a silver bullet, and can’t be used for every brand. Here are three problems I’ve identified with using share of search as a PR measurement tool:

1. Keywords. Just as with media listening tools, keywords are so important. The more uncommon the company or brand name typically is, the more accurate the data will be. If the keyword is taken out of context for the brand, or is too generic, it can skew the numbers and make PR attribution very difficult. Not everyone Googles ‘Apple’ to learn about computers.

2. More traffic = more problems. Large, global brands, especially those with an online platform with high volumes of traffic, make it very difficult to separate the new attributed traffic from regular users. Even if there is a peak in traffic following an announcement, a marginal increase will make it tricky to convince the boardroom of the value.

3. Targeting the right media. We’ve tested this for several brands, and if a news announcement isn’t targeting higher tier media (i.e. those with a high domain authority) and regular readers, it won’t show a tangible impact on search.

SoS can be an effective measure of PR impact, but only on a case-by-case basis. If used alongside other KPIs without thinking, it risks being just another vanity metric, thrown into the mix to try and wow a client without any real value. But if it's aligned to specific goals and objectives, and used with the right announcement, it can show real measurable impact.