“Do you ever go over games in your head? When you’re alone, I mean. Play all the way through them?” Beth Harmon, The Queen’s Gambit.
When I first started writing social listening queries in multiple languages, my colleague asked me “Have you seen “The Queen’s Gambit?”. It’ll help.
I hadn’t. So I did, and I got it.
Query writing is a bit like chess.
You have to keep going over the possibilities. Thinking about your strategy. Trying, failing, analysing what went wrong, and trying again.
We’ve been working with a global consumer goods brand recently, helping them build out queries across a range of different languages. As part of this, we’ve been sharing our learnings across the team and thought we’d share a few tips for any other wannabe grandmasters, trying to hone their multilingual query writing skills.
1. Word for word, like for like
Simply swapping out words and phrases for the destination language, is rough and ready but will get you somewhere (especially where the languages have similarities e.g. Spanish and Portuguese).
But this really is the bottom of the barrel.
You’ll always need to update the query according to the way people talk in each country. This can be particularly tricky for Spanish queries because it’s such a multi-dimensional language with a lot of different nuances to it and different expressions in every Spanish-speaking country.
As a general rule, a simple like for like translation of queries won’t work very well, our next tip explains why.
2. Coffee culture
The same word, when applied to different cultures can improve the relevancy of your results or rubbish them.
This means that you need to understand the nature of the word or phrase and the social and cultural context in which it’s used.
For example, in Portugal, coffee in the middle of the afternoon usually means there will be a snack with it (I like their style).
While in France, coffee tends to be drunk on its own.
When you’re looking to find people talking about eating, this makes a big difference.
Another point to keep in mind. You might know Spanish from Spain, but you’ll need to adapt to make it work in other Spanish speaking countries.
For example: in Chile, people often use the word “chiquillo” or “washito” for a child but in Argentina, they use “nene”. In Chile they use “papi” or “papito” for dad but in Argentina, the preferred words are “mi viejo” or “Viejo / vieijito”.
Completely different words that mean the exact same thing. You need local insight to ensure your queries are capturing everything they should.
Here’s a question. Are people using multiple languages on Instagram? It’s worth checking as it can really help you find the right conversations.
A couple of examples. We found using the English term ‘healthy’ helped us find more relevant conversation in French. The English was maintained whereas in other languages it wasn’t.
Another example is where the most natural local term has come to be associated with something else (e.g. in our project “bio” in French tended to relate to the political and (un)sustainable undertones of the organic market, so an English term is used instead (“organic”).
4. Exclusions and assumptions
Sometimes exclusions are not enough.
If you’re creating different queries for the same project and have been building out groups of keywords you might have to remove a few in order to increase relevancy. It’s important to know that every group might not work the same way for your different queries.
The same is true for your exclusions. These will need to be developed, tested and refined for each query. As a general rule, start with the broadest exclusions and build your way up to the more niche, query-specific ones.
5. Many hands make light work (Spanish, French equivalents?)
Do you have any international colleagues or friends?
Talk with them.
Sounds simple but it does help. We all use different expressions so it’s good to get the opinion of others, particularly in multi-language projects. They may see things differently.
Getting different perspectives on the topic of your query with people who speak the same language or live in the same country can be so helpful.
For example, a simple ‘hey guys, what healthy snacks can you think of at the top of your head’ can do miracles for your keywords.
Even genius Beth needed help from other chess players and friends.
6. Irrelevant comments?
In your journey to the perfect query, you might discover some comments that might not be relevant for the project but are still interesting. Write them down.
Share them with your clients or colleagues. This might help open up new opportunities to explore.
7. Document everything
Keep track of new things you’ve learned, what worked and what didn’t, nothing about query writing is wasted time.
There’s always a lesson if you pay attention. Just make sure every step has been properly written down for future reference. We use Basecamp.
8. Label every move
Write down all of the different iterations of your query, and make sure you also properly label your samples.
Why? Well, when trying to improve queries over several iterations, we can move further away from the target relevancy than closer (including going from slightly over target to way below).
This can be frustrating, but documenting all of the iterations will make it easier to go back to the on-target queries and save you a lot of time down the line.
9. “My tranquillity needs to be refurbished” Alma Wheatley, Beth’s adoptive mother
Query writing can take you down a bit of a rabbit hole (and drive you mad).
So, if you’ve been working on one query for days and find yourself stuck, move on to another query or a different task, take a walk, listen to some music and try again the next day.
Just whatever you do, and no matter how tempting it might be, don’t give up! You can always drop us a line you get stuck (firstname.lastname@example.org).