The Brexit Cream Tea Day Prolepsis

Kicking off

(Note that I am not responsible for external web site security or contents: click links or search Twitter feeds at your own risk.)

2016 was the second year in which a National Cream Tea Day was promoted. Given last year’s interest and subsequent hype, 24th June 2016 was always going to be much bigger, and was accompanied by a 12 hour version (9am-9pm) of the regular 8-9pm Friday Twitter #CreamTeaHour. Cream Tea Review produced a wonderfully silly audiovisual which explained what it was all about.

What wasn’t known a year ago, was that the night of 23rd June 2016 would see the count of a momentous UK referendum – possibly the most significant political event in the UK’s last 40 years – culminating in the final tally at around 7 o’clock on the morning of 24th June, showing that 52% of those voting wanted the UK to leave the European Union (EU) – the so-called ‘Brexit’. (See what a couple of Londoners at the Eurostar terminal did with scones a couple of days later!)

On a personal note (and I’m telling you so you can guage for yourself the extent to which you think I have distorted my analysis), I was absolutely gutted by the result, having voted passionately to remain in the EU. I was already lined up as a telephone guest on BBC Radio Devon soon after midday to talk about cream teas. I spent a couple of hours or so after breakfast lamenting the result and questioning my part in the upcoming radio programme, given how unimportant cream teas were feeling to me at that time. I decided to track Twitter’s #NationalCreamTeaDay to see what the different moods were, and how I might reflect on relevant issues later on air. From a superficial scan of tweets coming down my timeline, I began to gain several impressions. These framed my mood and I decided not to back out. But when I was interviewed on air (at 12:23pm), I admitted I was gutted, and was concerned about the possibility that the humble cream tea might become a dog whistle to a nasty sort of nationalism. I was struck how National Cream Tea Day was being used – certainly as an outlet for the despair of ‘remain’ voters, but perhaps also for a proxy celebration about the result (none that I saw up to that point seemed to be obviously crowing).

Social network map

All 7,317 tweets (including retweets) using #NationalCreamTeaDay on 24th June 2016 were collated by me (see the Data analysis methods section if you’re interested in the details), and the following map (fig. 1) was produced including all 14,445 links between tweeters, and to their various hashtags (NB: multiple parallel connections form denser lines):

Figure 1: Links between 7,317 tweeters (using #NationalCreamTeaDay) and their other hashtags on 24th June 2016

Network Overall 160625

Five clear communities emerged from the data (in descending order of significance):

 1) Over 27% of tweets (green – see fig. 2) linked around the core ‘drivers’ of #NationalCreamTeaDay: the Cream Tea Society (an exclusive marketing partnership of Rhoddas cream and Tiptree jam) and the underlying concept: #CreamTea.

Figure 2: The heart of the ‘drivers’ nexus of links

Network Drivers 160625

2) More than 13% of tweets (mauve in fig.3) belonged to a group of ‘passengers’ clustering around alternative issues of the day, predominantly comprising #Brexit, #EURefResults, #EURef, #BringYourDogToWorkDay, and #FridayFeeling.

Figure 3: Alternative issues mentioned

Network Passengers 160625

3) Nearly 11% of tweets are part of a Devon-associated nexus (pink – see fig. 4) including #CreamFirst (a reference to the difference between the way cream teas are prepared in Devon as compared to its cream tea rival Cornwall).

Figure 4: Heart of the ‘Devon’ nexus of links

Network Devon 160625

4) 9% of tweets comprised a very tight knit subsidiary network (in blue in fig. 1) focused on cream tea-related free deals and competitions.

5) Nearly 6% of tweets diffusely but identifiably emerged within a more national and tourist context, spearheaded by Tourism GB (yellow – see fig.5).

Figure 5: Heart of the diffuse ‘tourism’ nexus

Network TourismGB 160625


The first activity (a retweet) emerged at three minutes past midnight, and the last three minutes before midnight. 38 tweets had emerged by 6AM, a further 2,086 (29%) by midday, 4,266 (58%) in the afternoon, and the remaining 927 (13%) came in the evening. Half the tweets had been sent by 1:48pm.

Figure 6: Tweet build-up

Timeline NCTD 2016


Original tweeting entities (individuals, organisations and at least one bot!) numbered 2,361 (i.e. those who did not obviously retweet material – although some perhaps used a different entity, or made small modifications).

The most prolific tweeters of the day were @Roddas_Cream and @tiptree, who between them and through their joint marketing entity – @CreamTeaSociety – added 414 tweets and retweets between them (6% of the total). @DevonFoodLovers was next with 63, and @TourismTweetsGB generated 52. Of the 5,103 tweeting entities, another 14 managed 10 or more contributions during the day. The 18% (913 entities) who contributed more than once produced 43% of the offerings.

Ignoring the combined number of followers (undoubtedly comprising vast degrees of overlap), the average number per tweeting entity was 4,887 and the median was 621. This included 21 tweeters with over 100,000 followers, top amongst them being @NatGeoTravel’s 2.3+ million and @Telegraph’s 1.7+ million. Of those entities making 10 or more contributions during the day, the average follower list was 5,452 (median 3,154), ranging from @VisitDevon’s 32,644 down (but ignoring a small Tiptree offshoot) to one prolific retweeter’s 335.

Although 167 tweets were tagged with a non-English language, only 19 amongst them were not actually written in English. Additionally, five non-English language tweets were tagged as English.

The user origin country for 78% of tweets appeared to be (or include) the UK, 1% from the USA, 1% from an assortment of 21 other countries, and 20% listed no country of user origin (or a nonsense one).

An attempt was made to differentiate between tweets sent on behalf of a business in the form of marketing (generally when a product or service was specifically mentioned) and those simply used by an individual tweeting their own thoughts. But an early attempt to classify each tweet’s originator as business or private was abandoned in light of this more pragmatic differentiation.


Retweets made up 55% (4,018 to be exact) of all the traffic. Initial tweets comprised 39% of the total (2,826), and 473 of the tweets were replies (6%, excluding two replies to retweets).


Word cloud

A quick way to highlight the main hashtags used during the day, is to create a word cloud. Figure 7 shows the 26 populating the highest magnitude of hashtags (from 244 mentions down to 24 mentions), but excluding competition-driven ones which occupied their own world of motivations (#win, #comp, #freebiefriday, #competition, and #giveaway).

Figure 6: The top 26 non-competition hashtags used in association with #NationalCreamTeaDay on 24th June 2016



Ignoring the retweets, repeats, and competition replies, digging under the surface of the remaining 3,126 English language tweets sent by their 2,337 entities revealed a number of themes (often several in one tweet):


1,796 (57%) of these tweets could be put in the marketing basket, and came from 474 tweeting entities (20% of them). I stopped subdividing them after 517 examples of themes had been catalogued (many also promoted more than one element), and the numbers are neither complete, nor can they be considered strictly representative, but they give an approximate sense of the spread and intent of their marketers:

  • 51% were promoting a venue (e.g. a tea shop),
  • 22% were promoting a food or drink item (e.g. a type of jam),
  • 7% promoted something only loosely associated with a cream tea (e.g. a table to eat off),
  • 6% promoted a geographical area (e.g. a town),
  • 5% were promoting a charity,
  • 5% were promoting something ancilliary to the cream tea itself (e.g. teapots),
  • 5% related to administration for National Cream Tea Day marketing, not least for their half day #CreamTeaHour marathon.


Let’s quickly dispatch what appears to me to be a remarkably small number of irrelevant messages (perhaps because #NationalCreamTeaDay flared up so ephemerally). Amongst the 27 were handfuls of lewd, spam, trolling and nonsense messages, the strangest perhaps being from what appears to be a bot called @caffeine_ebook, rather confabulatiously announcing:

aussies if i give everyone still in eu knocked it looks fabulous #nationalcreamteaday pop their euros now trending in the scorpion


Four marketing and three non-marketing tweets made reference to ‘green’ issues, focusing either on local production or recycling of left overs.

Two marketing tweets and eight non-marketing ones referenced diets, tending, like @saylittlething, to diminish them:

Damn, never good for the diet…pass the clotted cream #NationalCreamTeaDay

Twelve marketing and 11 non-marketing tweets made reference to the weather – not unexpected given its role as a key to polite public conversation in the UK.

Celebrating cream teas

46% of non-marketing tweets  (606/1330) were clearly and directly celebrating cream teas (saying so or using an ‘!’ alongside #NationalCreamTeaDay). As @lulubelle2433 simply stated,

On another note… isn’t it #NationalCreamTeaDay? I’m happy to celebrate that. #nomnomnom #anyexcuse #watchthewaistline #devonorcornwall

Undoubtedly more implied it by not expressing negativity. A further 39 aspirationally celebrated, if postponed perhaps due to religion, like @sabdotzed,

Clicking on #NationalCreamTeaDay was not a good idea when fasting 😦

or a boyfriend issue like the one @grayst13 faced,

My bf went out to buy the bits for cream teas and came back with creme fraiche #NationalCreamTeaDay This is the worst thing thats happened

Traditional controversies

The old chestnut of the pronunciation of the word ‘scone’ reared its head again, but barely: nine times amongst non-marketers like @_Charl_Parker,

Seeing #NationalCreamTeaDay trending all day has made me really want a scone. (Pronounced like bone not gone.)

and three more by marketeers, keen perhaps to spice up their wares.

More typically controversial was the question of whether jam or cream should go on the scone first: by far the pre-eminent issue amongst 186 non-marketing tweets (6%) relating to etiquette, as journalled by @JueRobWilPo,

#NationalCreamTeaDay descends into bitter anarchy as a divided nation fights over whether the cream or jam goes on the scone first.

Fully 9% of marketing types got in on the action too, though often fishing from the sidelines like @PlymSciencePark:

It’s here at last! #NationalCreamTeaDay! Jam & cream or cream & jam? How to do you have yours? We’d love to see your pics!

The EU referendum

What takes this analysis beyond the mundane; what makes it of much interest to me at all really, is the impact of the so-called ‘Brexit’ vote on so-called ‘National Cream Tea Day’ (apologies for @CallumNovelli who dares call the whole thing into question – expletives removed):

So it’s now #NationalCreamTeaDay ? A few days after #NationalSelfieDay? Who the **** is making these “national” days? You *******

And I offer this with commiserations to the 3% of non-marketers like @bluelisa4 (who literally found out in the eleventh hour):

#NationalCreamTeaDay I only just found out, damn you #EURefResults taking all the limelight today!! 🙂

Celebrating Brexit

Stereotypically, @WillBlackWriter summed up what I suspect were many others expectations (prejudices?):

Yes Brexiters have a #NationalCreamTeaDay – while markets plunge and all the immigrants doing the hard work think Britain hates them

But in the chaos of the day on which the result was revealed, the biggest surprise to me has been the lack of celebration by members of the majority 52% who voted to leave the EU, not least in association with National Cream Tea Day. In all, I could find only 15 examples amongst non-marketing tweets, and some of them seem tenuous, intent made probable only by examining their writer’s wider public tweet trail. Among the more obvious, this from @Donna_R_E:

Fittingly, today is #nationalcreamteaday. We’ll be eating Victory Scones later this afternoon. #HappyIndependenceDayUK

Where were the rest? Are only miniscule numbers of pro-Brexit voters Twitter users, thereby skewing results? And if not tweeting, might they at least be found eating? I wondered where the marketing types stood: might many be hedging their bets; furtive scones for furtive celebrators? None clearly celebrated Brexit, which perhaps wasn’t so surprising, given the ease of alienating a large slice of potential customers. At least 11 of them recognised the import of the day but neutered it with a marketing message, like this from @siancothivale:

It’s #NationalCreamTeaDay-celebrate/commiserate #EUreferendum result with #happycows by buying British

Fifteen went a little further, appearing to go out of their way to emphasise British tradition – several even using the term ‘quintessential’ (rendering me rather queasy), such as this one from @mylegoman (emphasis theirs):

It’s #NationalCreamTeaDay today…. But where is YOUR favourite place to INDULDGE in QUINTESSENTIAL British Culture?

I wonder though how many marketeers were actually like dazed and discombobulated rabbits caught in onrushing headlights, as this one from our friends @CreamTeaSociety makes me imagine (still almost market-fresh at nearly 6pm, although punctuation is starting to waver):

Yes,a lovely British experience #NationalCreamTeaDay

A couple of tweeters at least took their conundrum to heart, like @rosemary_kelly_:

Feeling very sorry for these guys trying to promote clotted cream & #NationalCreamTeaDay. Not really in the mood…

A handful weren’t so charitable, at best suggesting some variant of naivety, e.g. @bouncingdan:

join the debate (almost entirely among brands) at the blissful island of ignorance here #NationalCreamTeaDay

but at worst implying dark forces at loose in the cream tea world, e.g. @ChrisTMarsden:

“Let them eat #NationalCreamTeaDay” brought to you by the type of #MiddleEngland #Brexiteers who have caused our current catastrophe

And @scytheanon was quite calculating in their depreciation of national Cream Tea Day:

#NationalCreamTeaDay is now only worth about 90% of what it was yesterday.

Other celebrating

What else was being celebrated in association with National Cream tea Day? Twelve marketers conjured up birthdays, dogs, memories, kissing, and even fairies, joined by 38 non-marketers. Of course, Friday is widely celebrated in the UK, and conjoined to 37 marketing and 39 non-marketing messages. But I couldn’t find much of interest to analyse here.


This is where one of the day’s most fascinating dichotomies took root.

Accentuate the negative

On the one hand, fully 53 (4%) of non-marketing tweets perceived a (variably) sinister connection between National Cream Tea Day and Brexit (none, as I’d expect, amongst marketeers). Most gentle were those wishing for one last kiss, some with imagery like @alison_hilton, others verbally like @JHJohnIsserlis:

#NationalCreamTeaDay sorry just not in the mood. To depressed after #EUreferendum would rather be celebrating with a continental breakfast

A few tasted bitter irony, like @StevieBuckley:

Poundland promoting #NationalCreamTeaDay is peak brexit

Whilst several couldn’t help but swear, their collagues found other ways to vent their anger, like @purelypoetry:

It’s #NationalCreamTeaDay today. You know, a good old British, independent dish. I’m gonna smash up a scone in protest. #Brexit.

A couple, such as @ unknownsock_zar, appeared to go so far as to call it racist:

It’s #NationalCreamTeaDay or as Percy would say racist tea

For several, the scone has at last gone down on the empire (e.g. @bluebmx71):

#NationalCreamTeaDay ruined for me after this vote, may as well put a blackshirt on and start goose stepping into town.

For others like @truekindhelpful, only dark humour remains:

On #NationalCreamTeaDay over 65 Brexiters all over England are saying to young people: “Thought you had a future, didn’t you? Look! Scone!”

Five or six, like @nasi_v, just felt alienated:

On a day like this, who in their right mind tweets about #NationalCreamTeaDay That makes me want 2 leave the UK more than any referendum

Perhaps @frankieisswell summed up the cognitive dissonance best:

When #NationalCreamTeaDay can’t make you feel proud to be British, something is terribly wrong #Brexit

Lastly, we had a couple of defiant gestures, like this from @stephanieee_hw:

Of course it’s #NationalCreamTeaDay today. Maybe Farage will choke on a scone.

Accentuate the positive

From 73 non-marketing tweets (5%) and 15 marketing ones, a different spin emerged. Amongst marketers, it might simply have been an indulgent ebullience, as suggested by @FingoMarketing’s:

Did you know that today is #NationalCreamTeaDay? It’s all so terribly British. We’re not complaining #FridayFeeling

More interesting though, were the plethora of non-marketing voices. Some were definitely on the celebratory edge, like this from @redbiochem , which incidentally suggests that the humble cream tea is still a novice in the nationalist stakes:

How British can you get #EUref results are announced on #NationalCreamTeaDay to become independent. All we need now is a GnT.

Others like @LeeTrewhela perhaps saw the gloves coming off:

Happy National (though probably not European) Cream Tea Day – @CreamTeaSociety #NationalCreamTeaDay

But there also came a surging, defensive undercurrent, snatching the cream tea back from the nasty orcs, to shine as the light of Eärendil in our dark place. As @JessicaYabbit (Galadriel?) puts it,

Finally, a reason to be proud to be British (ish). #nationalcreamteaday #scones

echoed by @lesleymckie (Samwise?):

The fact that #NationalCreamTeaDay is trending above Boris and Gove gives me some hope on a dark day.

But irony still found its way (even if unintentionally?) as discovered by @KarisTN:

Gosh, Britain leaving the EU. And on #NationalCreamTeaDay too.

And at least one tweet gave a stage to my cynicism about shire-myths, if only for a moment – honest (thanks @Nick_Corben):

Amongst the madness of today, #NationalCreamTeaDay starts trending.  This is why I love our country! #EURefResults #EUref #Eureferendum

although others, like @only_a_puppet, were much more witty than I ever could be,

UK instantly batten(berging) down the hatches and re-establishing sovereignty with #NationalCreamTeaDay. Take that, Europe. #EURefResults

Maybe there is a God after all, eh @HayleyWaily?

Thank heavens for #NationalCreamTeaDay. Love you Britain.


Clearly, a day like this was going to have outcomes, and here they are:


What’s not in doubt was the outpouring of crushed souls on National Cream Tea Day. 21% of non-marketing tweets said so, emphatically (and sometimes unrepeatably); two thirds of these explicitly blaming Brexit. (A cynic wouldn’t question why only 18 marketing tweets acknowledged any kind of disappointment at all today, and only 9 of them were specific about Brexit.) @JulesItsjules was one of the earliest (around 7.15am) to say it:

and possibly quite unable to properly appreciate the fact that it’s #NationalCreamTeaDay which is also quite sad

And taking it personally, @LauraPeta’s:

Oh sod off, Nige. It’s #NationalCreamTeaDay, don’t take that away from us too. #Euref

A couple, like @Lutherholdus, wanted a day of mourning:

#NationalCreamTeaDay is trending… Not today, we shouldn’t be joyful today. #WhatHaveTheyDone

or another bite (@darrened1):

We’re going to have to do this whole EU thingy again, it can’t be Independence Day it’s #NationalCreamTeaDay I don’t do multitasking!

Others lashed out, like @BritishCheese,

“@Roddas_Cream: Not just any Friday.. It’s #NationalCreamTeaDay” >> now that’s putting a gloss on it. #EURefResult

or engaged in displacement activities, like @madlinsudn:

Time to refocus or I will lose today to post-result wallowing and doom-mongering. Reset with tweets about #NationalCreamTeaDay.

One or two, like @LizaJaneG, went to their happy place:

We have Hobnobs in the office, apparently it’s #NationalCreamTeaDay & the sun is shining -who needs a PM or EU membership?? #WhatHaveWeDone

over-reached (@kellaree):

You fools have ruined #NationalCreamTeaDay as I now can no longer afford tea leaves. #EURefResults

or found themselves “tossed around by every cream of clotrine” (apologies to St. Paul) like @aclowderofcats:

I am too annoyed for this. Cream tea?! That’s what’s trending?! The country has come- ohh, actually, they do look good #NationalCreamTeaDay

One or two more sought to widen the fight, like @TJHoop82:

#EURefResults #NicolaSturgeon  is showing her true colours! #Independence!  #Blabla  #NationalCreamTeaDay

or even take on all-comers, like @philviles:

Idiots who #VotedLeave have triggered the #WhatHaveWeDone hashtag & today feels rather miserable. But at least it’s #NationalCreamTeaDay.

Were some perhaps in danger of rallying everyone against them (like bickering families suddenly united against an outsider)? I know I found myself suddenly bristling at @doubleshiny:

Do you know what Devon and Cornwall? You voted Leave so stick your #nationalcreamteaday and enjoy your EU funding wasteland

Of course, you’d expect to find some sarcasm (thanks @CherylMarkosky):

Hey, depressing post-Referendum news from estate agents. But hey, it’s #NationalCreamTeaDay. Feel better now?

and worse – puns (@Rachbythesea):

We’ve scone from the EU on #NationalCreamTeaDay 😂 Sorry, I’ll get my coat, don’t want to get myself into a jam.

Even a bit of hyperbolic sulking (@FrivolousMonsta) and I say this from within my own slough of despond:

Not sure of the timing of #NationalCreamTeaDay. Do enjoy yours whilst squatting on rubble & being kept warm by the fire of a burning skip.

But scattered throughout were the understated (a national trait perhaps?) like @MeganDownes_:

Not an ideal result, but at least it’s #NationalCreamTeaDay

and the pathologically grateful (another national trait?) like @fran_lowe94:

It’s #NationalCreamTeaDay, so can everyone stop talking about Brexit and pass the scones please and thank you.


Peeking out in 14 moments (and again, a cynic wouldn’t bat their eyelids at there being only two such marketing tweets), a sense of national division was being felt on National Cream Tea Day, like that felt by @poppetymoffitt:

The country torn asunder. But The Telegraph is celebrating #NationalCreamTeaDay. I think I am going to cry now.

Some faced it with a sense of despondent resignation (@BexEsmith):

#shocking…I think Britain says goodbye to prosperity and will be seen as ‘divided’….But, it’s #NationalCreamTeaDay

or like @BarrettTweets, proposed a duel involving scones at dawn:

It is #NationalCreamTeaDay today, so perhaps that will be the weapon of choice? To celebrate or commiserate. Providing WW3 doesn’t spoil it!

For others like @Damsonplums, it would come too late:

#NationalCreamTeaDay somewhat overshadowed. 60%/40% Leave round here. Feels like the Civil War with the Royalists sweeping across the shire…

Others like @alanpeart felt more like poking the wound:

Kids – go take your nan out for #nationalcreamteaday and while she’s spreading her marmalade ask her how she voted. #EURefResults

Still others like @GuiseKnightcott saw ominous portents:

Great, as if #Brexit wasn’t dividing enough, it’s #NationalCreamTeaDay. Now the West Country will split between JamFirst and CreamFirst.


Somewhat more patronising was the tone of an almost division-balancing number of non-marketing and marketing ‘unity’ tweets, like this from @CreamTeaSociety,

What better to bring the country together than #NationalCreamTeaDay ? #Brexit

or (with just a twist of bitter lemon), @1deeoned,

The nation might be divided after the #EURefResults  but #NationalCreamTeaDay has the power to bring us all back together. Minus Cameron

or (perhaps a little too unrealistically), @LeedsTK,

In other news, it’s #NationalCreamTeaDay! Proposal, in the spirit of inclusion, jam first on one half of scone & cream first on other.

Cream tea therapy

Shining quietly through it all, a softer light of hope was brought through the 10% of non-marketing tweets who, like @AudreyFleur, could transmit it, carried aloft by cream tea nightingales (and assisted by 2% of marketing tweets):

So #NationalCreamTeaDay has collided with #Brexit. The polar opposites for the British. The God’s of tea have sent comfort. #alwaystea

Maybe the gods walk among us, ponders @OwnFone:

We think that #NationalCreamTeaDay was invented just for this #anxiety ridden day. However any cakes will do! #mobile

A methodological approach was proferred by @rebekah_fox:

Take time to relax and enjoy a cream tea this afternoon. Sit down to gather your thoughts or just be in the moment #NationalCreamTeaDay

But most came disorderly if quietly, like @lexiexlambert:

On this…umm…’historic’ day it’s tea & cake to the rescue for #NationalCreamTeaDay . I would like ALL the scones please. ALL of them.

or perhaps not so quietly (@AstraBloom):

#NationalCreamTeaDay How lovely n British.Will guzzle pint of cream and stamp scones into floor. #nonottbeingstiffupperlip #notmyvote

and maybe with a bit too much commitment, like @TOMHS,

Not all doom & gloom as it’s #NationalCreamTeaDay – I’m going to jump into a vat of clotted cream.. Just leave me there..#Brexit #british

A rearguard though, like @jenloumeredith, thought the emperor wasn’t wearing any clothes,

There’s nothing a cuppa can’t solve. Apart from Brexit. #NationalCreamTeaDay

but the rest of us tried to ignore them.

Bigging it up

And so to the rallying cries. Of course the marketers wanted a shout – 15 of them at least, like @SalisburyHouses, proposing their offer was the ultimate:

Whatever the day brings, what is nicer than tea in the garden? #NationalCreamTeaDay @cooksandmakers @ReevetheBaker

How exciting! Perhaps some (maybe @HotelsCo?) really hadn’t noticed…

How exciting, today is #NationalCreamTeaDay Take a look at our #AfternoonTea menu & book now

But stiff upper lips prevailing, it was the 169 non-marketing (13%), overwhelmingly-non-Brexit-supporting tweets which showed the most resolve and determination. Keeping watch through the darkest hours, like @FishStock1 at 1.38AM, we found them:

Oh yes! Far more important than #EUref #CreamFirst #NationalCreamTeaDay

Poking politicians in the eye like @DawnDanger, we found them:

Better than the crappy politics #NationalCreamTeaDay

Back-pedalling furiously, like @nellydotmedia, we found them:

Cream Tea…the thing I love most in the world…..after my husband & children of course!! 😉 #NationalCreamTeaDay

Loving social media, like @paulfmuldoon, we found them:

You have to love Twitter. In the middle of the Brexit related trends is #NationalCreamTeaDay

Bristling Lady Bracknell-like (reminders of my late beloved sister) and like @hannahwebb90, we found them:

How dare the #EUref overshadow #NationalCreamTeaDay. Priorities people!

Admonishing the media like @DaBoi77, we found them:

What? No way! Ive been so bombarded with #EUref and not one mention of #NationalCreamTeaDay way more important sort it out #bbcnews 😡

Straight out of Reservoir Dogs (@_QueenDrag), we found them:

#NationalCreamTeaDay This has made my ******* day

Getting their priorities right, like @CodeNameTanya we found them:

#NationalCreamTeaDay this is important. I’m going to celebrate on Sunday instead. With my mum.

And with such spirit, damn it (wipes away solitary tear), @loobyhill1985 at least played the game:

The torture of not being able to open the jam #NationalCreamTeaDay

Organisers feedback

Following national cream tea day, I was very pleased to be able to conduct a phone interview with Rodda’s marketing manager, Belinda Shipp, heading the team responsible for promoting and monitoring #NationalCreamTeaDay for most of the day.

I began by asking her how the day went. “Brilliantly,” she replied, “absolutely brilliantly – exceeded our expectations.” She went  on, “it was a particularly difficult day from a social media point of view, being the Brexit results day, so to perform so well and be trending most of the day against that kind of back drop, I think we did exceedingly well.”

I asked her if she had stumbled onto any of the negativity in the day’s feed? “A little bit,” she admitted but, “national cream tea day was actually organised before we knew it was going to be Brexit day,“ so, “we decided to go ahead regardless.“

How did she feel about unpleasant tweets directs at the day’s organisers? “I don’t recall being particularly impacted by that, but we were working shifts. I may have just missed it.”

And what about the potential for associating national cream tea day with a future so-called ‘Independence day’ on the same date? “We haven’t set national cream day for next year yet… The last two have been the last Friday in June… We’ll see how things go I suppose, as things move forwards, but that has yet to be decided.”

The PR company involved have also released a short report about the day, highlighting their strategy and results. Note in particular that National Cream Tea Day last year was, “Developed as a PR and marketing tool… to initially launch the wider platform – The Cream Tea Society… to ensure both brands (Roddas and Tiptree – ed.) were seen as experts on all things cream tea and to develop a greater emotional connection to consumers.” They go on to say, “This year, National Cream Tea Day fell on the same day as Brexit but we were careful to position it as a quintessentially British comfort for many and a welcome alternative to the ongoing debate around the referendum… The Cream Tea Society organically reached a total of 322,465 people on Twitter with 199,807 personal accounts engaging with our content. Including our reach on Facebook (25,772), our total reach across social media was over 350K.”

So there you go.


The social network map proved fascinating, relatively simple to grasp, and will clearly vary wildly in future years. Whether National Cream Tea Day will have enough momentum by 2017 to do without coaxing and tickling remains to be seen. It seems more unlikely that it will soon become the ‘International Cream Tea Day’, given its strong UK and English focus thus far. Let’s just hope that there will be more actual conversation then (not merely retweets) than this time, although the surge of marketers who seemed to have discovered National Cream Tea Day by late afternoon, suggests that they will be far more ready next year to dominate with their megaphones.

The word cloud does a good job at picking out most of the key topics broached. But almost none of the hashtags really speak of the greater or lesser emotion and intensity lying behind any of them. This only emerges in a detailed reading allowing a moderately fine-grained thematic analysis (see fig 7 if you like data visualisations). Superficially, the day was dominated by marketing and cream tea celebrating tweets. Underneath though, and aside from a scattering of variably interesting flotsam and jetsam, it was immediately clear that the results of the previous day’s EU referendum were hugely pervasive. A perhaps tempting hypothesis was to presume that the cream tea, being a uniquely British (even a Devon & Cornwall) speciality, could and would be co-opted by ‘Brexiteers’. No doubt some did, but if so, then much more covertly than I, for one, had expected. Of course it triggered a not unreasonable wave of affection for a nationally beloved past time – the cream tea. But this proved vastly different from a nationalistic fervour. Far from enjoying or denouncing the association of cream teas with so called ‘Independence day’, the Twitter mood was overwhelmingly against endorsing such an association, preferring to see ‘Brexit’ as the deviant, National Cream Tea Day as innocent, and harking back to the halcyon days in which a cream tea and the EU lay down in peace like the proverbial lion and lamb. A huge wave of grief rolled out across cream tea land on June 24th, but it was soothed with a soft, clotted, creamy salve. In the end, the narrative has portrayed Brexit’s losers as the righteous guardians of the rites of cream tea; theirs to defend against an exultantly nationalist scalping by the winners.

Where do I stand now? I want to embrace the cream tea and own it with a pleasure tinged with sadness. But heaven forbid that National Cream Tea Day, or its commercial backers, allow it to associate with some kind of nationalistic, ‘Independence day’. If in future there is so much as a hint of a clash (of dates or ideologies), I shall be foremost on the barricades, in defence of the cream tea against the encroachment of a #NationalistCreamTeaDay agenda. As @rozryan has succinctly put it,

Good luck #NationalCreamTeaDay PR people…

If we lose, I’ll bow out with @jhurstj who found that,

My Cream Tea tastes of disappointment and resentment. #NationalCreamTeaDay

But though we leave the EU, I hope to stand proudly with @DNCanoa:

Sorry Brexiters, you can’t have today as your national Independence Day, because it’s #NationalCreamTeaDay

Figure 8: Expressions stairway – frequencies and connections between the main themes expressed by individual tweeters on National Cream Tea Day 2016

Theme cascade 160715

Figure 8 contains information from all 3,504 Expressions (see later for a definition) made on Twitter on 24th June 2016 using #NationalCreamTeaDay. 90% of tweeters used just one or two expressions during the day; only eight tweeters managed five or more.

  • The largest theme – Marketing – is represented in yellow to the right of the diagramme. It indicates how many connections were made by marketers to another theme. (e.g. Most did not connect – as shown in the bottom bar. But many did connect to the ‘Celebrating NCTD’ theme, whilst none were negative about NCTD.)
  • The next 15 most common themes are listed on the left, with the remaining eight aggregated at the top as ‘Minor themes’.
  • Each black bar shows how frequently a non-marketing-associated expression was made containing that particular theme, ordered from the most common ‘source’ theme (including self-referring themes) at the bottom to the least common ‘source theme’ at the top. (e.g. The largest group are expressions of the ‘Celebration of NCTD’ theme, whilst the smallest group shown are expressions of the ‘Celebration of Brexit’ theme.)
  • Arrows lead up and right, from a commoner theme to a rarer one, their width being relative to number of such expressions made. (e.g. The largest arrow indicates that the most frequent thematic link was between ‘Celebrating Brexit’ and ‘Etiquette’ – which is generally a proxy for the debate about whether cream or jam should be spread first on a scone. However, a lack of arrow means that no-one both ‘Celebrated UK culture’ and ‘Denigrated nationalism’ during the day.)
  • The arrows’ bases deplete the bar representing the commoner theme, with the leftover unused portion of the bar representing expressions that did not link to any other theme. (e.g. Less than half of those ‘Celebrating NCTD’ actually expressed another theme during the day, but almost everbody expressing ‘Disappointment with Brexit’ wanted to say something else too.)
  • Arrows do not indicate a narrative direction – they merely indicate the direction from commoner to rarer themes. (e.g. A significant number of people who mentioned ‘Disappointment with Brexit’ went on talk about getting ‘Therapy in NCTD’ for it, but many who linked ‘Therapy in NCTD’ to an ‘Unspecified disappointment’ – probably Brexit given the circumnstances – tended to run the narrative direction from disappointment back to therapy rather than the direction the arrow might imply.)


Editing tweets

Where quoting tweets above, I have removed most @references, since they can misleadingly connect people to others’ opinions. I have also removed links because they tend to distract, may not be safe, and some contain elements which some readers may find offensive. Where tweeters have used a line break, I have tended to remove it and to substitute a full stop if needed, just to improve the formatting of the quote, but I have otherwise left spelling and punctuation as per the original. Lastly, I found that some symbols and emojis weren’t transferable from Twitter into Google docs, and have removed any odd characters which were automatically substituted.


I am not aware of any third party influence on my analysis (for example lobbying or money), which is entirely mine.

Mention has already been made of my pro-Remain position and disappointment at the vote, although I generally prefer participatory to representative democracy and will not support attempts by elected individuals to undermine the will of a majority referendum.

I doubt I influenced the content of #NationalCreamTeaDay significantly (and I chose not to advertise my book or website using #NationalCreamTeaDay), but:

  • a few days in advance, I was asked about how to enhance the Devon-related Twitter impact on National Cream Tea Day, and suggested to the enquirer that people might like to join the #CreamTeaHour discussion that day between 8-9pm (I didn’t join in myself).
  • I also replied , “Why?” to two unclear tweets on the day (without public replies).
  • I added my own tweet (with one like, but no replies or retweets) at 11:34AM saying,

It’s #MontyPython-esque: “We’re leaving the EU.” “Quick, eat a cream tea.” What shall I say on the radio later? #NationalCreamTeaDay

(Thanks to @twayward who’s tweet half an hour earlier must have subliminally primed me.)

  • I was briefly interviewed about cream teas on BBC Radio Devon at 12:23pm on the day

Data analysis methods

Using Google Docs online, I first loaded a template called TAGS v6.1. I then ran a search using it for all tweets incorporating #NationalCreamTeaDay and covering the 24 hours of 24th June 2016 (British Summer Time). (NB: Given the way the Twitter Search API works, this may not be reproducible for the same time period if attempted more than 7-9 days later. NNB: The template also notes that it is possible that the API misses some tweets if it does not think they are relevant.) I then downloaded the .xls version of the data scrape for analysis using Microsoft Excel 2010 (MSExcel) offline. The raw data contained 7377 tweets, of which 60 (0.8%) exact duplicates were removed, leaving 7,317 for analysis.

A .csv version of the data was rejigged using the open source OpenRefine tool and guidelines from School Of Data. (But retweets are sometimes truncated in the scraping process, so I created the ‘mentions’ column by using the extraction process for # and @ on a version of the ‘entities_str’ column in which # and @ had been substituted for “text”:” and “screen_name”:” respectively):

The Social Network Analysis was made using the open source Gephi 0.9.1 tool. Using edge weights and a resolution of 2.5, a Modularity calculation was run revealing a Modularity score of 0.755, and 469 communities of which 5 contained more than 5% of tweets. (NB: four links to an orphan # were removed as they would have led nowhere in the actual Twittersphere, and #NationalCreamTeaDay was removed since that was anyway the underlying search term and left me with the remaining links between these tweets.)

The Timeline was devised using MSExcel.

The analysis of Users was done with a pivot table in MSExcel.

The Word Cloud was generated using the online tool at WordItOut.

Themes were extracted using MSExcel:

  1. As well as removing retweets and obvious competition replies, repeat tweets were screened and deleted if they appeared to originate from similar entities (e.g. the second of a pair of identical tweets sent hours apart by @ParentingWT was removed, and 13 of 14 identically worded tweets –although with different links – from @wow247 affiliate entities sent within one hour were removed).
  2. Links were not reviewed unless the tweet was otherwise unintelligible (several hundred).
  3. “#NationalCreamTeaDay” was ignored in a tweet when classifying themes, unless referred to specifically (e.g. Happy #NationalCreamTeaDay), or accompanied by an ‘!’ (assumed to be celebrating it unless context suggested otherwise), or the only text (when a celebratory motivation was assumed for having sent the tweet at all).
  4. Themes were prospectively created, by iteratively condensing and recondensing, wording and rewording, each tweet in their time order. The earliest ones underwent most modification, producing an increasingly stable set of recurring ideas, classificatory phrases, and informal rules for handling divergences. A fairly coarse grained approach was adopted early on in order to limit the burgeoning of new categories to a manageable level (hopefully aiming for less than 50 so as to aid manual classification of so many tweets). By the end, 37 themes were listed, although 6 of those starting with ‘promoting’ were no longer collected after the umbrella term ‘marketing’ was introduced after 517 examples had been extracted. One of these 37 themes (‘homemade’) proved recurrent but didn’t really have any explanatory power and wasn’t analysed in the end.

Themes were then manually mapped using Microsoft Publisher as Expressions.

  1. Firstly, themes were concatenated with commas separating them (to obtain a single column of themes), then
  2. using OpenRefine, they were split back down but this time into multiple rows and exported back as a CSV file.
  3. Back in Excel, all columns were removed except the theme and the user ID, and duplicate rows were then removed, allowing each user tweeting entity one bite at any particular theme during the day. All rows not containing one of the 24 principle themes were also removed (mainly the ‘excuse to eat’ or ‘homemade’ themes). This left 3,504 unique expressions (combinations of a tweeting entity and a theme).
  4. Using a pivot table, theme frequency was ranked from 1-24 and an equivalent label added to each expression in a new end column. Expressions were then sorted first against user ID and second against theme rank. A formula was used in a new column to copy the next ranked theme for each user (for the 935 users with multiple expressions, or to self-refer the theme for the 1,353 users with only one entry), in order to create a catalogue of notional ‘Source’ and ‘Target’ rows. After deleting the 935 surplus rows, I was left with a list of 2,569 expression ‘edges’.(Where a user had more than one expression, they would have a final line which had no counterpart since it had been copied to the row above, but was not listed as self-referring.)
  5. These could have been easily imported and displayed in various ways in Gephi. (Change the column titles to ‘Source’ and ‘Target’, and to allow edges to scale, instead import a non-duplicated list of edges, but count and enter the numbers for each edge as a ‘Weight’ column.) However, I preferred to manually construct the cascade-type data visualisation in fig. 8.