Marketing: How to Measure Purchase Intent Using Social Listening Guide

By Phillip Agnew on March 20th 2015

During our recent UK masterclass, clients and staff alike were treated to a whole host of thought-provoking content.

The presentations from some Brandwatch superusers took us through the unique ways clients are using social data, including a big-name client’s use of Vizia and Check One Two’s use of social to educate millions about the threats of testicular cancer.

#feeling nuts banner

Some presentations demonstrated the truly unique insights users can glean from Brandwatch Analytics, specifically purchase intent language on social.

Mentions and engagement for your brand can be a good indicator of an effective social media strategy, but they don’t always equate to revenue or profit.


The importance of purchase intent

Intent to purchase is the conversation that happens online when a user expresses an actual desire for a product rather than generally talking about a brand or their product.

It might include a Tweet from a person expressing their desire to buy a product, a post about wanting to pre-order something before its release or a blog update indicating an eagerness to see a newly-released movie.

To help broaden our understanding of how this works, presenters at the Masterclass took us though how monitoring purchase intent has made a difference for one global brand.


H&M celebrity sponsorship

In Brandwatch’s recent retail report, social data was used to analyze H&M – the second largest fashion retailer in the world.

Brandwatch Analytics tracked H&M as they ran four major celebrity endorsement campaigns. The purpose of the research was to not just analyze which celebrity caused the greatest buzz, but which generated the most purchase intention mentions.

purchase intent

After collecting and analyzing the data it became clear that David Beckham drove the highest volume of conversation for the H&M brand. However, Beyonce provoked more purchase intent conversation.

Through social listening, H&M were able to measure the value of their endorsement in a whole new way. Previously David Beckham would have been seen as the best ‘value’ endorsement for the retail brand due to sheer magnitude of conversation generated by him and his undies.


You might like

Advanced Social Listening: Intent to Purchase Conversation & ROI

Read the Article

Now it’s clear that in terms of monetary value added to the business, Beyonce clearly comes out on top.

Brands like H&M now have the opportunity tie conversation with sales and create bespoke campaigns based on whether they wish to drive brand awareness or sales.

Screen Shot 2015-03-19 at 9.27.14 AM


How did they do it?

Fortunately, the Masterclass event made an effort to match presentations on success stories with step-by-step how-tos, meaning clients could put into action many of the insights they had heard.

Ross Barrett gave us a walk-through demonstration of how to set up purchase intent Queries within the Brandwatch Analytics platform.

An intent Query is a way to find mentions by people wanting to take an action; whether it’s buying something, going somewhere or doing something. They all start with personal pronouns; I OR Id OR “I’d” OR we OR we’, as seen in the example below.

personal pronouns purchase intent query

Next, add in action terms, like need OR want OR looking OR going OR booking.

At this stage you’ll also want to bring in the NEAR/xf operator. The NEAR states that the two words must be in proximity to one another and the letter f after the number indicates that the second word must follow the first in that order.

action purchase intent query

 

Finally, bring in context with the subject of the action – in this case holiday OR getaway OR “city break” OR cruise, for example.

Obviously, this is where you edit the Query to match the product of interest to you. This could be anything from a restaurant booking to a SasS platform.

context - intent to purchase query

 

And there you have it. A purchase intent Query showing exactly how many people are talking about buying your product.

Try out purchase intent monitoring for yourself by using the Query below. Simply replace [BRAND] with your brand name.

 

( ((I OR “I am” OR “I’m” OR Im OR really) NEAR/0f (((want OR wanna OR need OR looking OR “interested in” OR tempt* OR “Thinking about” OR “thinking of” OR ((cannot OR “can’t” OR cant) NEAR/0f wait) OR considering OR “feel like” OR “feeling like”) NEAR/2f (buy* OR purchas* OR get* OR order*)) NEAR/7f [BRAND]))
OR (((“I will” OR “I’ll” OR Ill OR “I’m” OR Im OR ((I OR “I am” OR “I’m” OR Im) NEAR/1f (“going to” OR gona OR gonna))) NEAR/0f (buy* OR purchas* OR get* OR order*)) NEAR/7f [BRAND]) )

We’ve got a huge backlog of Query strings set up to monitor a whole host of different conversations, including purchase intent, so feel free to get in touch if you would like to try it for your own brand.



Brandwatch Analytics

Discover purchase intent conversations by harnessing the power of Brandwatch

Find out more

Phillip Agnew

@p_agnew

Phillip is Product Marketing Manager here at Brandwatch, informing clients and others about our latest platform updates.

  • Georgiana A. Murariu

    This is very useful, however because it is so broad/inclusive, it will result in a massive pool of data that will take a long time to process. Any tips for how to tackle those initial stages of processing/sifting through content? I know this is valid for any other string/search query but just wondered if you had some tips. I.e. once you get your data in with the intent of purchase, would you for instance make rules/tags/categories that contain something like ‘don’t’ or ‘no’ or ‘not today’ to separate between immediate intent to purchase and eventual intent/long-term planning? Thanks :)

  • Chris McCormick

    Hi Georgiana,
    So first of all, if you want to segment the data further i’d have the following suggestions:
    – Like you mention, we typically write rules that simulate the customer journey.. ‘any recommendations?’ – > ‘thinking about buying’ -> ‘going to buy’ -> ‘in store/on the site’ -> delivery/opening -> past tense conversation “i bought/have been using” etc
    Typically we’d create each rule to capture conversation specific to the brand/industry and then assign sub categories of a ‘customer journey’ parent category to split them all out.
    If you want even further granularity you could create tags for each of your products/ranges/brands and cross reference the buying cycle of each tag.
    From there you could identify pipeline drop-off points, estimate demand for newly launched products by benchmarking against existing ones, cross reference sentiment in early stages with volume/ratio of purchase, the possibilities are endless!
    Similar can be done to replicate net promoter score using personal pronouns and emotive keywords to segment the data.
    We have the community forum open now if you have any more questions your brandwatch login will allow you to ask the experts in there!
    Have a good new year.