We focus on recommendations, because advocacy drives sales

Online buzz is full of recommendations, both positive and negative. As a major global brand you already know that these recommendations are immensely powerful in driving sales through influencing consumer decisions.

Because online advocacy is so important in driving sales, we have developed a dedicated WaveMetrix methodology for tracking recommendations.

What is a recommendation?

Not all online discussion contains a recommendation. A lot of buzz is just conversation, but occasionally recommendations occur, and when they do they can be impactful.

The key factors which differentiate recommendations from other posts are:

• 1. Statements of merit. Recommendations are always concerned with the relative quality of a brand, product or service. The statements can be positive or negative. Consumers use words like “the best” or “the worst” or refer to specific use cases such as “Inglourious Basterds is the best film for livening up an otherwise boring evening”
• 2. Overall conclusions. Recommendations differ from other posts in that they account for all the positive and negative factors a poster perceives in reaching an overall conclusion. Posters use words like “Inglourious Basterds is the best movie out at the moment” or “I recommend Inglourious Basterds because it is so much more entertaining than any other movie currently out”

There are two types of recommendations, explicit and implicit. Both types of recommendation share the two essential features described above (statements of merit and overall conclusions). Where they differ is in the strength of advocacy of the poster:

• 1. Explicit recommendations. A recommendation is explicit when the poster is imparting advice and trying to change the choice or behaviour of the reader. Examples include “I recommend Inglourious Basterds”, “If I were you I would choose Inglourious Basterds” or “Don’t go and see Inglourious Basterds, whatever you do”
• 2. Implicit recommendations. Implicit recommendations are statements of merit and overall conclusions just like explicit recommendations. However, they do not involve intent on the part of the poster to influence the choice or behaviour of the reader. Examples include “I think the best movie in cinemas is Inglourious Basterds”, or “I’ve seen a few films in the last month and Inglourious Basterds is easily the best”

Why do we distinguish between explicit and implicit recommendations? Both are important in influencing consumer choice, but advocacy is stronger in the case of explicit recommendations. We use this distinction to make sure that we not only capture all recommendations (both implicit and explicit), but also differences in the strength of advocacy.

How does WaveMetrix identify recommendations?

Advocacy cannot be identified or coded by automated social media monitoring solutions.

At WaveMetrix we can identify recommendations because all buzz is read and coded by human analysts. This ensures we understand the content of posts and can identify which are recommendations and which are not.

Using human analysts means we can also identify and classify the reasons why people make certain recommendations and not others. We call these reasons 'advocacy drivers'. Analysing online recommendations by advocacy drivers gives brands a clear picture of why people recommend their products.

Inglorious Basterds: advocacy by type

Graph based on buzz around Quentin Tarantino’s Inglourious Basterds generated in August 2009. 32% of total buzz involved a recommendation, either implicit or explicit; but the majority of recommendations were implicit.

Measuring the impact of recommendations

As described in Principle 4: Engagement not all posts are created equal. Some are read by millions and others hardly at all.

The same applies to recommendations. It is therefore essential to measure which recommendations are read the most and to allocate most weight to these. WaveMetrix uses the same impressions methodology as is used for normal opinions to measure the impact of recommendations. See here.

Using impressions we can measure which recommendations have most influence, but also which sites they are on and which posters make them. We can also examine advocacy buzz alongside other data – for instance, sales figures.

Top five movies UK 21^st - 27^th August 2009: advocacy buzz vs. sales

Note: for these recently released movies, advocacy buzz levels and UK sales figures correlate closely - suggesting measuring online advocacy buzz is a good predictor of box office success

Inglourious Basterds: UK advocacy buzz and sentiment by topic

We commonly split posters into four advocate types:

• 1. Active advocates. These people are making positive and explicit recommendations about the brand, its products or services. The brand should seek to understand what enthuses advocacy drivers. They are natural candidates for seeding viral campaigns
• 2. Active detractors. By making explicit recommendations against the brand these posters often (but not always) have a substantial negative impact on sales. The brand needs to start by understanding what concerns them - identifying key negative advocacy drivers. Further actions should be based on this insight and often include enhancements to the next version of a product or sharing additional information with the posters in question
• 3. Passive advocates. Passive advocates are brand followers who are making recommendations, but whose recommendations are implicit rather than explicit. Brands should look for differences between active and passive advocates to see if there are opportunities to make passive advocates more vocal
• 4. Passive detractors. These posters are just like active detractors, but less vocal. Brands need to review advocacy drivers to understand why negative recommendations are being made and develop a plan to address key negative issues

Inglourious Basterds: advocacy matrix