Who knew Alexis Mac Allister was so damn popular?
Not me. To be 100% honest I didn’t even always remember his surname was two separate words until some time mid-November.
I know he seems like a nice guy – that’s evident in the genuine joy both from him and his team-mates as he returns to Brighton post World Cup triumph.
I also know he doesn’t seem the type of player who is going to angle for a move away – though as an Arsenal fan I wouldn’t be averse to him following Leandro Trossard up the M23.
But particularly popular? More popular than Bukayo Saka, Phil Foden, Neymar and Marcus Rashford? More popular than Luka Modric, Antoine Griezmann and Harry Maguire combined? And this during the World Cup? And this among people in the UK?
No, I didn’t know that.
However, thanks to Google Trends and a lot of time spent gathering data and manipulating spreadsheets I do now. I thought I’d leave it seven pars to mention spreadsheets, hopefully your interest is piqued enough to continue.
I’ll keep the methodology short – though I have covered it previously. I will also point out right now that I know that over the course of the World Cup Mac Allister was not more popular than Bukayo Saka in the UK (although it is a lot closer than you might expect).
What I really want to do is chat about why this matters, how popularity is actually more predictable than it might seem and how this can help drive content planning. If you know what people are going to search for – and it’s not the obvious stuff like Lionel Messi – you can prioritise resource and boost site traffic. Just as importantly, you can give people what they want.
Some Words on the Method
(skip this bit if you know the methodology)
I wanted to capture Google Trends data for every player at the recent World Cup whose team made the quarter-finals – that’s squad players too. On completion, I could confidently name all eight third-choice keepers of the quarter-finalists, though don’t test me now.
In addition to this, I added a few other nations of interest – Belgium, Spain, Germany and Wales. Three European heavyweights who did unexpectedly poorly, and the nation that was my home for four years at university. No prizes for guessing…
Google Trends allow you to compare five terms – or players here – at a time. Set the date range, the country you want to analyse and away you go. However, don’t do a new set of five each time, if you do that your data is useless as trends will always rank something in any comparison as 100.
If you compare Matty Cash v Lucas Digne for the past year, Cash will peak at 100. If you then compare Lio Messi and Ronaldo, it is Messi who will peak at 100, with Ronaldo at 59.
But, you can’t take Cash’s 100 alongside Ronaldo’s 59. Run a comparison of Matty Cash and Ronaldo and Ronaldo peaks at 100. Cash tops out at five. I suspect I’m the first person ever to run that comparison by the way.
You need a control. My control was Christian Pulisic. It needs to be someone who is mid-ranking, and during the World Cup he was just that, much like all the other Chelsea attacking mids who are technically gifted, but you wonder quite what they contribute. Mason Mount, Hakim Ziyech and Kai Havertz would also have made perfectly reasonable controls – so too whoever they have signed this week.
Pulisic had a decent peak when he scored for the US in a game that has since slipped from my memory, but there will be many players higher – and an awful lot lower.
Pulisic then appears in every comparison, it is him and four others. If he is the peak search item, topping at 100, the data for the other players can be entered as is. If, instead he peaks at say 50, then all the data needs doubling for fairness – the peak in this one might be Gareth Bale at 100. Which becomes 200.
I removed all scores below 5 – I could explain why at length but trust me instead. It is for the best.
You end up with a big table of data that you can see via this link.
https://docs.google.com/spreadsheets/d/1_SzDs0cya0LqTBHQ_MNw4u9M1yt_oIRMWoDc_OMEFI0/edit?usp=sharing
I collated the data for both the UK and then also global searches. RIP actual work deadlines during that week.
You can see the ebbs and flows for each player, it is fairly obvious when a player might peak, but the actual values less obvious. Harry Kane’s google search soars like a ballooned penalty on the day of England v France, but his peak at 527 (five and a bit Pulisics) is much less than Raheem Sterling’s peak during the World Cup of 909, just a sixth of Messi’s peak and roughly double that of Angel Di Maria’s peak score (unsurprisingly on the day of the final).
And then there is the global data to factor in – so let’s get to what this means, and its use.
Where the data is useful -and where it isn’t
I included the link to the data above, but for those of you who skipped that stage, let’s ctrl-v it here too.
https://docs.google.com/spreadsheets/d/1_SzDs0cya0LqTBHQ_MNw4u9M1yt_oIRMWoDc_OMEFI0/edit?usp=sharing
There you go. Hopefully I set permissions correctly. No editing it…
Let’s start by answering a common query. What use is it knowing how search interest in players ebbs and flow? Kane is obviously searched for come World Cup time, nobody is going to search for Portugal’s reserve left back and if Lionel Messi scores twice in the World Cup final people are going to bung his name into Google.
Also, even if someone temporarily surges, what can you do with it? That young guy who came in for Ronaldo and scored a hat-trick – Goncalo Ramos – people searched for him during the match, but a day later? Not so much? Two days later – back to almost nothing. Unless you have a feature on him ready to go, what can you do – and if you have that feature ready it is fair to assume you have wasted a lot of effort also producing features should they be needed for Ante Budimir, Jonas Hofmann and Jordan Morris. (Not yet well-known Chelsea players at the time of writing)
Instead, the data is less useful for being reactive during an event, it is of more use to understand how people search, what they search for, and what this means for future events – football, sporting in general and other.
The lessons I take from it, and I reckon there are plenty more to be found, are:
1. The predictable interest in the unknown is valuable
Returning to Mac Allister, this is a player who is not unknown, but he is on the fringes. Pre-World Cup, he was just one of many Premier League players heading off, and one from one of the smaller clubs too.
For roughly the first week of the World Cup, there was very little interest in him at all in terms of UK searches. Why would there be?
However, think about how this changes. Interest in general in Argentina grows, especially after they lose their opening game and then recover to make the knockout stages.
They then play out a dramatic knockout game against the Netherlands on route to advancing to the final. By this stage, Mac Allister is a regular in the team. He is a young midfielder playing at the heart of the Argentina side, and playing very well too – he plays in England but has rarely made the headlines there, he is at a club that attracts limited media attention.
If he has a good game in the final, if Argentina win, then there is almost bound to be interest – and this is also true of sex-act-simulating, troll-in chief Emi Martinez by the way, only interest in Martinez has already been spotted and profiles written.
Mac Allister does have a good game, a very good game and of course the game itself is a classic. Interest I Messi and Mbappe is huge, but Mac Allister registers 370 – for context this is very close to peak interest seen in Jude Bellingham in England during the tournament, and just think how much attention he got.
Mac Allister also remained heavily searched for the following day. Anyone who mentioned him in their live text of the final, or had him featured prominently in the match report, or wrote some sort of profile – or better still had one ready to go – would be in line for a major boost in site traffic.
Key to this is balance. There are many players who COULD become rich sources of search traffic – if we think solely of people searching from England, then anyone scoring a winning goal against England is a contender, and so too is anyone getting sent off for England, or perhaps missing a penalty in a shootout. You could come up with a list of 100 players.
Instead, it is opportunities where a high degree of search is almost inevitable – in the World Cup final, this was any England-based player likely to feature in the game, and after that it is trying to find that sweet spot of those players where search interest will be highest, but competition from other news sources lowest. Mac Allister was always going to stand out.
Had Belgium made the final, Leandro Trossard would have been a better candidate than De Bruyne or Lukaku – less well known to start with, lower competition from others, but also still likely to do something noteworthy in the game itself.
This approach can be used to forward plan in any event. The Wimbledon draw, cast your eye forward. Who will Andy Murray meet in round three after the inevitability of two gruelling five-set wins. It’s probably someone seeded about 26 or 27, they won’t be well known at all. However, when they beat Murray after four tough sets -and also during this match – their name will be going into a lot of Google searches…
2. Some search is universal, some is local
I will move on from Mac Allister in a bit, but he is useful here too.
You might assume that as an Argentine, search interest in him will be even higher globally than it would be in the UK< but you’d be wrong. Well, obviously more people searched for him globally, but he stood out from the crowd much less.
In the UK, his peak was 7th highest of all players, and his overall search total – adding up scores across the tournament was an impressive 15th. This is higher than Richarlison for instance, the scorer of one of the great World Cup goals. Across the tournament, there was more than three times more search interest in Mac Allister than there was in Huge Lloris.
Globally, however, his peak was 24th and his overall total 30th – he is below Lloris and he is almost level with Enzo Fernandez. Let’s talk about Fernandez.
Enzo Fernandez also played in the World Cup final, he was even named the tournament’s best young player after the match – is that Golden Balls, or was that a show with Jasper Carrot in which game theory suggests you just say ‘steal’ and then deal with the consequences afterwards?
Fernandez ranked 165th for peak search interest in the UK, and 206th for total search interest – this is below Nick Pope, it is below Toby Alderweireld, and why was anyone searching for him?
But, globally, he is ranked 31st for peak search and 33rd for total search, his total is near as dammit the same as Mac Allister’s, whereas in the UK for every time his name was searched, Mac Allister’s was about 80 times!
That is staggering.
The major difference is obvious – Mac Allister plays in England and this one aspect alone led to a lot of local search. If, come the next World Cup, Fernandez is playing in England, his search levels would likely surge too – he may even replace Pulisic as a control subject, after all he’ll probably be at Stamford Bridge, or at least out on loan from there.
Ben White is another example. Like Raheem Sterling, he left the World Cup early, though for different reasons, and Sterling of course returned. Leaving the camp led to UK-based interest in White, his peak value was 236 – two and third Pulisics and his total of just over 1,000 was the 13th highest of all.
Globally, though, he ranked 96th both for peak search and total – in the nicest possible way, people heading to Google outside the UK weren’t overly fussed about England’s third (fourth? Fifth? Choice right-back heading home early. After all, they might rightly think, the rest of the squad won’t be too far behind.
Sterling, meanwhile, ranked 3rd for peak interest and 7th for total in the UK, but 23rd for the former and 34th for the latter globally. Globally, he is behind Enzo Fernandez for both, yet more than 150 places ahead in the UK.
In short, in 1990s music terms, some footballers are Bros – big in the UK, irrelevant globally, others are Garth Brooks and just a few are Madonna, Messi for example – just two letters away from a much-more common comparison for him.
Also, this is the best I could come up with on putting Bros into giphy. I’m not sure if it’s about the brothers Goss?
Anyway, naturally these geographical differences need factoring in when running UK and international editions of indexes – just because a story is big in the UK, it might be of very little interest globally. Equally, others will be making the news elsewhere yet getting little attention in the UK. Vinicius Jr ranked 21st globally for overall search, yet 75th in the UK – Karim Benzema was fifth – FIFTH – globally for total search despite not even getting on the pitch.
Let’s just say, all other things being equal, if you had to prioritise writing a Ben White or a Karim Benzema story mid-World Cup, you’d be making a clear decision about whether you want UK traffic of global. As I said, all other things being equal.
3. Some interest is super brief
Goncalo Ramos’ ranking for peak search was 16th in the UK, one place above Ben White and one below Olivier Giroud, but their overall search was very different. Ramos got less than half the search overall of the other two in the UK.
Ramos came in – that in itself newsworthy as he replaced Ronaldo – then had the audacity to score a hat-trick. That’s going to make you searchable.
However, two days later his search was back to very low levels and, while there was some interest in him for Portugal’s quarter-final it wasn’t particularly noteworthy. Certainly not something to be impacting editorial planning.
White probably wouldn’t have attracted any search interest but for his need to head home, but that then became a story that was covered for a few days, there were two days of very high search and several more with decent levels.
Giroud meanwhile got pretty high levels of search four times during the tournament – I’m not entirely sure why to be honest, bar him being subbed off in the final – and did he score against England? I’ve blocked that one out too.,
For a quickly-emerging, and just-as-quickly vanishing topic such as Ramos, there isn’t actually a great need to do things any differently editorially. If you are running a live commentary of the game, it is going to mention him heavily, so too will the match report. You could perhaps bash out a quick profile, but then you’ll probably make follow-up features about him in any case.
Perhaps the one think to note is to be quick, search for topics like these has often dropped greatly inside just an hour or two. Put your follow-up feature live the following morning and people’s search intents have move on.
If the surge in search is linked to a match then it is generally only going to remain at most until the next match – and in an event such as the World Cup that is only ever a few hours away.
I wanted to put a Kaiser Soze image here – and like that, he is gone, but I think that might be frowned upon nowadays.
4. The big names are REALLY BIG
It would surprise nobody to learn that Lionel Messi had the highest total search both in the UK and globally. It won’t be jaw-dropping to learn that Ronaldo, Neymar and Kylian Mbappe rounded out the top four, albeit in different orders in the UK and globally.
What is noteworthy is just how many searches the big names generate even on days when they’re not doing much. Rest days, days before their team has even played, or days after their team has been eliminated.
We can look at the minimum search value, and for most players it is zero – this isn’t zero actual search but it might as well be. It’s less than 1% of peak search in Christian Pulisic, and who’s getting out of bed for that?
In the UK, the minimum level of search for Messi was 67 and for Ronaldo it was 117. The least search Ronaldo received on any day from a day before the World Cup to a day after was still more than Antoine Griezmann received on any day – and yet he seemed to be chatted about loads. I even found myself Googling his name at one stage, and not for the making of this.
Globally, the least Messi registered for a day was 200 – two Pulisics – and for Ronaldo it was a frankly-unbelievable 333, three and a third Pulisics, and this Pulisic at his most popular, and he’s a pretty popular player.
Looking at worldwide search, Ronaldo’s lowest daily score of 333 would rank him seventh among players’ peak scores for days. No England player’s peak came close to Ronaldo’s bottom (so to speak) – Harry Kane highest with 233.
When Kane blazed the penalty over v France, his interest surged – but still only 2/3 of the level Ronaldo’s was at on a day he did nothing, and by nothing I mean even less than the day he sat out and pretended to be happy as Ramos netted a hat-trick.
We therefore see huge differences in overall search. Globally, Messi and Ronaldo were both searched for around 20 times more often than Kane and 50 times more often than Sterling. Neymar was searched for 30 times more often than Marcus Rashford – and 45 times more than Youssef En-Nesyri who, lest we forget, helped fire Morocco to the semi-final. He bested Ronaldo, just not in search interest.
There are players who emerge to become of interest, there are the big names, and then there are the HUGE names. These will dominate search pretty much regardless of how their team does.
5. If in doubt, stay local
Everyone wants to rank for Messi and Ronaldo but, let’s face it, everyone will write about them and so it won’t be easy. There is a lot more search, but it is a lot harder to rank.
Is it better to rank on page 7 for Messi or page for 1 for Mac Allister the day there is interest in him – not Messi-esque interest, but interest nonetheless.
I’d take the latter any day.
But what of the quite big names, the ones who we might be able to predict high interest in when planning coverage.
Good advice is just to say local.
Messi, Ronaldo, Neymar and Mbappe were the top four for total search in the UK and globally, but after that the picture varies greatly.
Globally, the highest-ranked England player for total search was Kane in 13th.
In the UK, positions 5 to 11 were entirely taken by England players. It reads like a Garth Crooks midfield. Hopefully, I got Brooks and Crooks the right way round in this post, though it’s debatable whether a Garth Crooks album would be worse than a Brooks one, and at least Brooks might consider playing a defensive mid.
If we compare and contrast, we see – with their UK ranking and then global.
Bellingham (5th, 14th)
Kane (6th,13th)
Sterling (7th, 34th)
Foden (8th, 28th)
Saka (9th, 23rd)
Grealish (10th, 56th)
Rashford (11th, 52nd)
Then we see
White (13th, 96th)
Maguire (16th, 45th)
And
Mount (20th, 102nd). I mean, Mount is probably about 14th just among Chelsea attacking mids globally.
If you’re doing a World Cup round-up, or a quiz, or a picture gallery, or World Cup gossip, or any other potential Pulitzer prize winner, then in the UK these are names to bung in if you want that search boost. They might be in your SEO kicker (shudder).
And, if you’re in the US go for Pulisic, and Weah and Adams, and if you’re in Australia chose. Err. You get the idea. Tim Cahill and John Aloisi? Are they still going?
6. The low numbers are just as interesting
Studying the search interest in performers across any event can tell us as much about those who don’t attract significant interest as those who do. This prevents the wasting of effort, it aids prioritisation.
Most players in a tournament such as the World Cup won’t stand out search-wise at any stage – probably over 90% fall into this category. After all, only a few teams really capture the imagination and within each of those only a handful of players.
However, that still leaves plenty.
From the 2022 World Cup, it is not hard to think of players who maybe might have trended. Croatia made the semis, so what about Modric, or Spurs’ Ivan Perisic? Or what about Wales keeper Wayne Hennessey, the first player sent off at Qatar World Cup, or Cody Gakpo who enjoyed a prolific start to the tournament for the Netherlands?
There was interest in all these players, but nothing remarkable. There was a small spike in search for Hennessey on the day he was sent off, but this quickly passed. Gakpo and Perisic had levels of search that would not demand changing any plans for and Modric was fairly well searched for – a few Pulisics worth – but you could expect this of a well-known player captaining his side to the semi-finals. The aim of work such as this is to find new insights, not to reaffirm that Luka Modric is quite popular.
Morocco are interesting – semi-finalists but no player ranking particularly high for total search in the UK. Globally, there is more interest and the wonderful Achraf Hakimi ranking 11th – this likely to be purely from the quality of his performances. There are two other Morocco players in the top 20 – goalkeeper ‘Bono’ and Hakim Ziyech.
En Nesryi didn’t cut through, ranking just 78th globally. This I think shows that being quite good in a team that does well is not enough, for someone to really gain interest there have to be a few aspects to the narrative.
For Mac Allister this was the fact he plays in England, at a smaller Premier League club, is young, relatively unknown – or at least most people won’t know a huge amount about him – and is playing in a World Cup final.
A player like Gakpo, however, is just quite a good centre forward scoring in group games that are, let’s face it, forgotten by most people two hours later when the next match starts. Had he scored a winner against Argentina it might have been different.
Oh, and speaking of that, Wout Weghorst did score twice against Argentina (who saw that coming?) but he only ranked 73rd globally and 68th in the UK. He scored two late goals, but his country still lost.
Who cares? Or, who cares enough for that to motivate them to search his name in Google?
7. The need to be selective
To finish, and building on the previous point, we learn that actually we can be selective.
Purely in search terms, there won’t be that many names who really become of interest during an event, outside of the obvious people we could all predict pre-event.
This knowledge allows time to plan and focus, you are not looking for 30 lesser-known players to target – at least in search terms – instead it is perhaps a handful.
There is no need to jump all over every half-chance, you don’t need to get your star writer to bash out a quick profile of someone who has just scored a brace for Ecuador v Poland in the opening Group B match.
The fact someone has done something interesting is just the starting point, after that the question is would I, as a sports fan, bother to put their name into Google? Really? What am I hoping to find out?
I might want to know who Mac Allister joined Brighton from, how old he is, or his market value. But, En Nesryi scoring in the quarter-finals for Morocco? Well, after the game has finished we are straight into England v France and the moment is gone.
What are you doing as a user then, watching the next quarter-final, or heading to Google?
At Wimbledon, the unknown Brit beating the equally unknown number 31 seed in round one won’t actually lead to particularly high search in the UK. There is another match straight after – several matches, there are other talking points.
But, in 2022 you might have looked at the draw, seen who Serena was set to face in rounds one and two and thought – hmm, one of these might well be Serena’s final game. This opponent, ranked 117, is completely unknown if they beat Serena on centre court and retire her. Well, people will search for her won’t they… (and they will – but be quick).
What Else Can We Learn?
I’m hoping you can help me out here. I have other ideas, but this is getting lengthy isn’t it. We can compare across sports, how do the top golfers compare to the top tennis players? We can see the growth in women’s sport, we can see that for cricket franchise stars such as Jos Buttler and Sam Curran are often far more searched for than a Test great such as Joe Root. This, in part, why the first two go for millions in the IPL, and Root continually remains unsold for all his skill.
And, I could go on.
If this is your sort of thing, put ideas in the comments or drop me a line if you want to chat through ideas, or this field of work.
How else can trends data be used for event planning?
If This Has Been Of Use
This took bloody ages to make – especially the research stage and there is much more that could be said about the data.
The data is also available to all, head over and interrogate it, download a copy if you want to play with offline.
If it has been of use and you’re feeling kind, please share it, or drop me a comment, or kiss the screen or something.
Better yet, I can train individuals or teams on the process so you can use trends (which is 100% free of course) to really help with future planning. Or, I can do it for you. Planning for an event but not sure which players or people cut through best, or who might emerge as that lightning in a bottle gold?
Leave it with me…
Well don’t. Get in touch, agree a project, and then leave it with me.
