Google and its ilk rank by authority. Our platform — earth.ai — ranks by engagement, the driver of behaviour. Therefore, although you can use the same queries as you use in Search Engines, you may find that the tips here will help you more quickly construct queries which get you closer to the heart of the engaging aspects of your topics and concerns. These few simple rules can help you quickly achieve more focus and deeper insight.
1. Target Issues and Concerns
The best queries aren’t guesses — they’re issues or concerns you know are keeping you or your client awake at night.
The standard process for orientation, is to run a full Landscape taxonomy (all the things that keep the client awake at night), and have the platform show us what matters, and how to effect what matters.
But, in response to a ‘show me something’ question, a natural tendency is to quickly run a Narrative Analysis on ‘something’.
So ok, what exactly should that ‘something’ be?
If you don’t know your client’s pain, you can look at issues within their industry/sector — because we report engagement, including those topics can help. It’s also a safe bet that brand and business jointly spur engagement. CSR and supply-chain issues increasingly impact brand value and business performance. What was the last such scare to rattle their industry?
What’s the best query — e.g. is it SMB cloud, SMB cloud issues or business problems with cloud? Ask yourself, what’s the vernacular? Find out how real people talk about this topic. A quick Google (or Bing!) search will give you some idea of which language is prevalent and appropriate.
2. Think Like a Journalist (or Steal from One)
What would be the ‘headline’ summary of such topics? Mouse in Hamburger Scandal will probably capture engagement better than BurgerCo reviews supply chain rodent contamination. If the client is already in the news, have a look at what language actual journalists have used — chances are that other voices will pick that up and use it.
3. Get Local. Get Specific. Get Emotional.
‘Single word’ queries usually fail to capture engagement. Even on a biggie like Climate Change you are better localising, specifying and adding in emotive language. What about Climate Change adaptation Australia or Climate Change Murray River farmers or even Climate Change economic threat. Engaging narratives are usually about something changing in some way, or acting in some manner. Not Apple but Apple profits fall for example, or not Cadbury Chocolate but Cadbury Chocolate Africa child slavery perhaps.
Likewise: often using overly formal language will take you down a too formal path: the language businesses and brands use to describe their worlds is often very distant from the language used by the real world to describe them, their world, and its concerns. Just because a brewer issues press releases about localised problematic alcohol consumption, doesn’t mean that the narrative isn’t Sydney drunk mob rampage by the time it hits Twitter.
And again, engagement is often at a local level, spurred by personal or community experience. Often, using too high a level will steer you way from the real meat: e.g. Australian retail may serve as a good proxy for overall economic health, whereas Sydney retail may better capture street-level engagement.
4. Find Analogies
Which of the client’s competitors has recently experienced issues, or have had engaging events occur in relation to them? A query on those can help demonstrate ‘what it would look like if this — heaven forfend — happened to you.’ Forewarned is forearmed, right?
5. Iterate
Once you’ve run an initial Narrative Analysis, refinement is often as simple as adding another term to the run or — as you might do with a search engine — using inverted commas to tell the platform to treat what’s in the commas as a single term.
i.e. do you want University of Sydney (there lots of universities in Sydney, so that query will help quantify how much each is contributing to that overall narrative) vs ‘University of Sydney’ (which will return an analysis of engagement around that specific entity.
Also, have a thoughtful look at the Emotional Response and Topics and Themes metrics for your initial run — if, for example, there is a high degree of Expectation, and a strong association between certain topics not included in your original query, add them in, to explore the angles which are really generating engagement. This could, for example, take you from an initial Trump to Trump widow insult hate as a deeper dive into the heart of that specific narrative.