Let's talk about Project Omega, one of the most important milestones in the history of data science, which revolutionised the supermarket trade.
You know it better as Tesco Clubcard
WARNING: Thread contains an actual swordfight /1
In the early 90s, Tesco was in BIG trouble. They had gotten complacent. They were being squeezed by growing discount chains below, and had sunk millions into failed attempts to reposition themselves as a 'quality' brand and dislodge Sainsbury's
Enter Terry Leahy.
Leahy was a marketing man whose star had been rising in Tesco. In 1992 he became Marketing Director at just 36.
But Leahy had a dirty secret. He wanted to do something utterly heretical to the Tesco board. Something that was a forbidden phrase.
Leahy wanted a Loyalty Scheme.
Loyalty Schemes were utterly verbotten in Tesco because of one thing: Green Shield Stamps.
GS Stamps allowed people to collect tokens and cash them in for other goods. Founder Jack Cohen had personally signed Tesco up to them in 1963. They had made it the grocery behemoth it was
But by the late 70s Green Shield Stamps were crippling Tesco. They weren't enticing new customers, but paying for stuff existing ones got via stamps was costing Tesco £20m a year (£275m in today's money).
But Jack Cohen and the old guard were emotionally attached to the Stamps.
It took years for the younger execs to get rid of Green Shield Stamps at Tesco. There was practically a palace coup, led by then Marketing Director Ian MacLaurin in order to do it. Careers were ruined, friendships betrayed. Cohen forced to step back.
Tempers got so heated that ceremonial swords were ripped off the boardroom wall in Cheshunt and there was a LITERAL SWORD FIGHT over Green Shield Stamps.
In the end MacLaurin won. They were ditched. The money plowed into Operation Checkout, kicking off Tesco's second golden age.
So yeah. Loyalty schemes? Not something you talked about in Tesco, especially as MacLaurin was now CEO.
But Leahy had been watching the Co-Op dividend model for years. And on a visit to Bury he'd seen them trialling a magnetic strip system, collecting user till data too.
And what Leahy realised was that everyone was thinking about it wrong. Yes, TRADITIONAL loyalty schemes were of marginal benefit.
But what if you could combine that with getting LEGIT customer purchase data as well? That would be decision-making gold, surely?
On 18 Oct 1993, sign up tables appeared at Tescos in Sidcup, Wisbech and Dartford Tunnel. Shoppers were offered a plastic card, 6 free cups of tea, and a thank you letter from the manager. In return, Tesco wanted their name, address, age and family size.
This was Project Omega.
Project Omega was Leahy's covert loyalty scheme pilot (Omega was a furniture store over the road from Tesco's head office).
It was run by a young marketing employee called Grant Harrison who, it turned out, had just done an MBA thesis on loyalty marketing.
Harrison would prove critical to Clubcard's success. He'd poured over why loyalty schemes went wrong in his thesis, and from the start decided that Clubcard need to be:
- Simple
- Delayed rewards (people don't 'see' discounts)
- Unrestricted. Your points should convert to cash.
For six months Omega was just Harrison, with help from a small direct marketing firm (Evans Hunt Scott) based above a carpet shop on the Edgeware Road.
Over time, this grew to a small team working out of Riverside House, a near-abandoned Tesco office scheduled for demolition
It was one of that team, Jane Lacey, who came up with something that turned out to be critical:
Clubcard had to feel like it rewarded EXISTING loyalty. Not an incentive for FUTURE loyalty. It had to feel like a 'thank you' for customers to sign up and engage.
She was right.
The other big question was what to collect. Here, Tesco discovered with horror that their rival Safeway was ahead of the game. They were already piloting a scheme that allowed them to collect info on EVERYTHING that was in the shopping cart, item by item, called ABC.
Tescos didn't have the tech to do that. BUT... Siemens Nixdorf confirmed they could make Clubcard piggyback off the Tescos Electronic Point of Sale (EPOC) tech they'd already installed.
This meant limited data though. Would it be enough?
Enter dunnhumby.
dunnhumby were a small (at that point) firm founded by Edwina Dunn and Clive Humby. Harrison had seen Humby, a mathematician by background, talk about analysing shopping data at a conference.
dunnhumby were drafted in to see what use could come out of the data gathered.
dunnhumby quickly realised something critical:
The EPOC data was more useful than Tesco thought. Yes it didn't have specific items. But it grouped spend by DEPARTMENT ("nursery", "alcohol", "baked" etc).
That was enough to start drawing useful conclusions.
Project Omega was go
By early 1994 it was clear to Leahy, Harrison et al. that they were onto something. In the trial stores customers were signing up like mad. Soon £6 out of every £10 spent was from Clubcard holders. Lacey had been right: customers saw it as a 'thank you' and a perk, not a discount
And the data they were getting was game-changing. They could see individual shopper behaviour in store and behaviour over time, and react to it within a week.
But one thing worried them: Clubcard equated to 1% of profits given back to customers. Would they spend more in return?
Leahy was, for want of another phrase, utterly bricking it about the first voucher send out. Essentially Tesco was printing money, and hoping customers would spend that on something new or extra, not just their regular shop.
Turned out they did. Sales went up 4%
(Managing that first voucher send turned out to be a nightmare btw, once they realised that they were effectively sending £5m in cash through the post. Tescos had to bring in bank-level security measures for the vouchers AND posting. I could do a whole thread on voucher design)
And so at the end of 1994, Leahy, Harrison and Humby finally did the thing they'd been dreading:
They presented their loyalty scheme results to the Tesco board. A board composed of the people who'd couped Cohen and killed Green Shields.
Including CEO Ian MacLaurin.
They ran through their powerpoint, showed all the sales data. Explained the ROI, then finished by going through key things they'd learned about the customers via the data.
MacLaurin and the board watched this, in silence.
"Any questions?"
More silence.
Then MacLaurin spoke.
"What scares me about this," MacLaurin said, "is that you know more about my customers in 3 months, than I know in 30 years."
He told them they had 12 weeks to get Clubcard into EVERY Tesco in the country. Oh, and if any of the opposition launched a scheme first? Gameover.
Again, don't ask me why I know/retain this stuff. I've no fucking idea.
Everyone else is watching Game of Thrones or something and I'm at the back all: "can't watch. am immersed in the incestuous and backstabby history of consumer loyalty schemes"
(Prices shifted to current prices to make it easier to understand)
So at Xmas they knew they needed to stock (and push) a "premium" wine range, perhaps accompanied with a discount offer in the pre-xmas voucher send out to people in the right purchase/item history bracket.
Okay bonus 'History of Tesco Clubcard' story because why not.
At launch Clubcard points only kicked in when your cart value was over £10.
This led to Tescos getting complaints from pensioners, so they lowered it for them.
And then they thought... "hmmm what about students?" /1
So Tesco wondered whether they should add a student category with a lower starting point as well.
So they decided to run a few focus groups of students to see how well that would go down.
Now Tesco focus groups in the nineties were a relatively genteel affair. Because the company legit valued people giving up their time. They'd be invited into a room with beer, wine and a buffet to chat to a moderator, while execs watched through one-way glass next door.
Where we left off the small team working on a loyalty scheme in Tesco, heresy within the firm, had revealed their semi-secret Clubcard pilot to the board
GOOD NEWS: They weren't getting fired
OH SHIT NEWS: They now had 12 weeks to get it in EVERY Tesco /1
Hindsight: CEO MacLaurin signing off on Clubcard shouldn't have been a surprise. Tesco had been floating on the edge of bankruptcy. They needed to innovate to survive.
That was also why they needed it NOW and MacLaurin's believed they had to be first-mover. Else it wouldn't work
This left the Clubcard team with a problem though. The pilot had peaked at 12 stores. Scaling to NATIONAL was a whole different ball game.
With the biggest question being: how the HELL do you process and store that much customer transaction data in the frikkin 90s?!
I don't think there's any sector that, for the most basic things, likes to shout:
"But i've not been traiiiiiiiined by someone!"
Louder than Higher Education.
I do wonder how some people make it through normal life.
"Can we go to the park dad?"
"Sorry son, not been trained."
The irony being that the sector is also legit bad at actually providing training for the REALLY complex stuff where it's perfectly reasonable for a user to want training.
It's all very, very, silly.
You end up with people refusing to do things that are functionally the same as stuff they do at home until they've gone on a 1 day training course delivered by 2 trainers.
Whilst also being asked to do complex shit that they should get a week's training with no human help at all