Expertise

What to remember about the practical use of data

15 minutes

By Virginie Chollois, Agile Marketing Manager senior at Lucky cart. 2022/10/24

We remind you of the title of our speech: Driving your business with first party data, myth or reality ?

Indeed, we are currently right at the heart of the matter. Data allows us to analyse, to see what the eye cannot see, to predict shoppers’ consumption and therefore has an immediate impact on business.
Once the subject has been established, let’s go into a little more detail.
With us, on this October 4th 2022, Olivier Dauvers as moderator, surrounded by Stéphane Smolarek (Intermarché), Jérémy Hoy (GroupM Commerce), Jérôme Hamrit (In The Memory) and Romain Charles (CEO Lucky cart). Nothing but good people!

Here, we will try to summarize the 1h30 of free, constructive and enriching discussions that we were lucky enough to have and to set the stage for our next exchanges (and yes, there will be others!).

A point on context, activity and challenges around data

First, Olivier gave an overview of the retail sector and discussed the role of data in a difficult, hard-fought market (even more so than before with the advent of new players), which must constantly reinvent itself in order to continue to move forward, to ‘modernise’.
Commercial times are tough and will continue to be so. With this in mind, what are the solutions available to us, to you? The answer seems simple: to be the customer’s favorite retailer. But how do you put it into practice? Here are some of the answers through 4 preference levers:

  1. the offer
  2. the shopping experience
  3. image/price
  4. the brand/customer link.

OK, but how can data bring meaning, richness and therefore impact?

  • for the offer, by diversifying the assortments, by bringing a different product catalog from the shop next door. Bringing a personalized, varied offer adapted to my needs.
  • + 4. for the shopping experience and the brand/customer link, it is imperative to personalize the messages sent to customers, by targeting communications more. “Talk to me about me, I’m only interested in that”.
  • for image/price with yield management ? Realistic? Utopian? Technically possible at least! Even if we are still a long way from it and the French are probably not ready for this 360° change in price/product usage. To be continued…

Once this introduction has been smoothly conducted by Olivier Dauvers himself, let us begin our exchange.

Part 1: The state of play of data.

To get an enlightened view and various answers, we interviewed : Jérôme Hamrit.

For him, the state of play is clear: “many retailers have laid the foundations: they collect, clean, store and archive their data.
The level of maturity in the use of data is quite different between our five major retailers, and so are the use cases.
The organization around this new project is not always optimal and for it to work at its best, it must be based around a key couple: the retailer and the data professionals, those who understand it better than anyone else, those who handle it with skills and respect for the rules.
We are not there yet, the paradigm shift and its implications are not always easy to implement, the road is still long, but we are on the right track.” 

A clear state of play, which says a lot about the road ahead…

The paradigm shift mentioned by Jérôme Hamrit is also and above all a cultural issue, an acceptance of a change of gear in the shopper relationship which is moving from the mass market to a 1to1, personalized, personified relationship, the ultimate goal. But today, understanding what is being done, the normal path that data takes, its use, goes through key stages, it brings a little more intelligence and methods to operational staff to think differently, save time and imagine activations differently. Not easy!

To support this approach, Jeremy Hoy shares his view:

“The positive value in this major change is that data allows us to initiate the transformation so necessary today. The volumes of data are there, aggregated in real time, and allow media players, among others, to measure each approach, each campaign, each message and to have a more holistic and non-media centric approach.
Yes, the model is renewing itself, it is even reinventing itself because it is driven by globality. Data does not only serve the food industry but much more widely. This allows everyone to feed off it, to be inspired by it and to have a clear target vision.”

Pure-players are structurally more advanced than traditional retailers, but the best cases can nevertheless be inspiring for all!

On the basis of this state of play and the road ahead, we discussed the internal organizations and the changes in the professions that data requires.

Part 2: organization, the offer…:

This is a major subject if these changes are to be carried out in the best possible way, with finesse and fluidity. For Jérôme Hamrit, “not all retailers have the necessary business expertise within their organizations to play the role of gatekeeper, or pivot.
It is clear that current organizations require a change in structure to put in place teams of ‘knowers’ associated with specialized service providers. One of the challenges will therefore be to recruit profiles that know how to identify and qualify needs and rely on the right business partners to manage these high-stakes strategic projects.

This organizational point is very clear for the large-scale food retailing sector and transformations will still take place to optimize the whole, in the months to come.

The other key point when a player wants to use data as a major business lever is the offer.

Who better than a retailer’s representative to give an answer on this point: Stéphane Smolarek
“the offer that adapts to the seasons, to the needs of shoppers… It is technically possible and yet my retailer is not yet implementing it. It’s a long project, which will come to fruition if all goes well in mid-2023.”


In addition to the personalization of the offer, data also enables prediction or more precisely predictive modeling, which is a set of methods for collecting and analyzing transactional data, so as to interpret it and ultimately deduce consumer behavior in the near future.
These are proven probabilities, built on a tangible, documented basis, which, in all likelihood and given the amount of data analyzed, can be considered plausible, precise and probable results.

Olivier Dauvers repeats the information given by Mourad Bensadik, Carrefour’s e-commerce Director at the Ateliers du Drive: “We are able to predict 70% of a basket from 3 orders placed by a customer. Personalization is a major element in e-commerce.” Is this realistic?
We asked Jérôme Hamrit: “Yes, it’s totally realistic because of the predictability of e-commerce, due to the complexity of the processes… This prediction is a great loyalty lever for retailers.”

Let’s now focus on a key issue: the personalization of the offer.

Are there any obstacles to shopper consent, is it difficult to obtain, and what image does this process convey?

For Stéphane Smolarek “acceptance is essential”.

For Jéröme Hamrit “the personalization of the offer and promotion is still in its infancy. Because today, we are still in the vast majority of cases on mass promotional investments, in shops and therefore not personalized.
When you think about the digital issues, the loss online and the huge sums of money that are distributed to shoppers who don’t ask for it, it makes you think!
We need to see how quickly the personalization of promotions is possible, but the players are a little ‘skittish’ about changing the rules and actions that have already been tried and tested by everyone.  The process will be slow to implement, but the client is ready to accept it.
The problem is diving into the unknown vs. what you already know. You know the returns from mass campaigns and substituting them for a lever (e-retail) that is not yet as powerful (in terms of number of customers reached), the switch is difficult.
It’s a slow transition that we’ll have to make together.

For Romain Charles “it’s also a problem of human nature… We talk about brands but it’s exactly the same for industrial brands. For years the rule of the leaflets has been ‘you deliver 100, you will sell 70 and of these 70 sold, 50% will be incremental, the rest is cannibalisation’. These are the same figures, for years.

Finally, Jérôme Hamrit adds to his previous comment: “You know that 10% of your customers are pure fans of promotion. They hunt for the best deals, and may even change brands to get the best possible prices.
If you take this into account, and if we are talking about data, you can identify them, contact them, offer them the best deals and make sure they don’t go elsewhere. Once you’ve done that, you can ease up on the pressure level for the other 90%. And then you start another discussion, and frankly there’s more to do!

To summarize the beginning of the discussion, we have two major projects:

  1. the first data usage project, the construction of the offer.
  2.  the second area, personalisation.

Then comes a more media-oriented topic of data use.

Can retail ecosystems become media platforms?
Olivier Dauvers says: “a FMCG brand, in its media arbitration, had yesterday: TF1, M6… Today, with the same media arbitration, it still has TF1, M6 but also the Carrefour ecosystem, Infinity… Retailers will become the new media gateway for brands.”

The question was asked to Jérémy Hoy, who is Managing Director of GroupM Commerce at GroupM France: “That’s it exactly! Because today, the brand’s site is captive in its media inventory, in its capacity to propose a potential advertising investment.
On the other hand, since our topic of the day is data, we can use shoppers’ data with the broadcasting ecosystems, whether they be TV, billboards or others…
Data is the opium for brands and media players, but we need to take into account all this knowledge that retailers have and make it available, so that it can be used according to the real need. 

For Jérôme Hamrit, on the same question: “the direction is the right one. The important point is cookieless, which is coming quickly, and what will happen to measurement, to the ability to measure without it?
In concrete terms, when we know how to track an individual and we have the right tracking kpi’s such as campaign ROI, we are therefore able to track the buying behavior of shoppers both off and online.


Part 3: The Magic Wand

Olivier Dauvers continues his questions and here is the one we were all waiting for: if you had a magic wand, what would be the use case or a perspective in terms of unconstrained data, without limits. What would you like to achieve one day?

Stéphane Smolarek is the first to answer this question and to take his Harry Potter wand: “consolidate the data of food and non-food retailers, while respecting the rules and the consumer, in order to be able to offer him the right product at the right place. This is to take the mental load off him when he shops. There is a complementarity between the sources of purchase, the brands that can be done.”

Next, Jérôme Hamrit and his magic wand: “data should allow for end-to-end integration in the supply chain.
The detection of a sale or a breakage at a given point in a shop immediately takes the whole chain back to the manufacturer who produces, to help with follow-up, reduce waste, and manage the safety stock that wastes a lot of time, money and raw materials.
If the ecosystems and data were integrated, they could save time and money for everyone involved.
Technology from the checkout to raw material management.” 

The question continues with Jérémy Hoy’s magic wand: “by going beyond the divisions of traditional brands and cross categories, we can have a data room with all the food, service and non-food data… A large data bank. With a view to rationality, to be able to have the various multi-category purchasing paths, in the service of a more measured and proven marketing/advertising pressure”.

And to conclude, Romain Charles uses his magic wand “by reviewing promotion. We are now sure of mass promotion, with relatively low transformation rates.
Today, the business model of leaflets (which will disappear, by the way) is to communicate massively on promotions with relatively low transformation. But the results are still there and this marketing practice has existed for a long time.
The end of leaflets is a real opportunity because if you really want to improve your conversion rate, you have to be sure that the offer made to your customer is the right one, that the brand proposed is the right one, that the format proposed is the right one. Today, leaflets can be a barrier to purchase for some people because even if the brand interests you, the format pushes you to store for 6 months!
So if you are able to adopt the right format, to push the right offer according to their buying habits, it’s a safe bet that the conversion rate will be better.

Not bad for our sorcerer’s apprentices, eh? 

The final line with a Q&A session:

Question on the reconciliation of data and distribution channels. What role(s) for brands? – Thanks to Florence from Datagram.

Stéphane Smolarek: “Some brands will necessarily be easier to convince because they are a concatenation of brands.
And then, at some point, it will be necessary to test and perhaps a voluntary brand associated with a voluntary retailer on data sharing will be simpler to carry out than a roll-out to the 5 retailers with more than 20 million customers”.

Romain Charles: “The role of brands is crucial. In this tripartite role between the retailer, the brand and the business partner who can support them. The role of business knowledge is extremely important and this business knowledge also comes from the brands.
Beyond the brands, it is the role of the people in the brands that is major.

Finally, for Jérôme Hamrit: “data has the merit of being objective, factual and fairly indisputable. It creates a platform for collaboration between brands, retailers and partners. They use the same data, work with the same kpi’s and have the same analyses”.

Question on impulse buying: how can data be used to support impulse buying? – Thanks to Stéphanie from Mondelez International.

Stéphane Smolarek : “This is a real subject! Data can help in this area, but it is still not addressed by retailers today.
Pushing the right product, with an ultra-relevant offer to encourage transformation and not feel oppressed”.

The use of data can play a role as a prescriber, recommending alternative products that will change the intention to buy.

Stéphane Smolarek : “If we go to the end of the data, we personalize the messages and recommendations according to their buying habits.
On the other hand, if the shopper is not sensitive to this use of data, we will have to question ourselves and propose alternatives.

Before concluding, a big project in the making : personalization.

If a retailer can get to know its customers even better, its alternatives will have even more value perceived by the customer. We are opening up a major data project.

Romain Charles says on this subject: “Digital and data allow us to do things in a much easier way than in-store: measure efficiently and quickly.
And if you use A/B test, you can measure the action even more effectively and get insights that are not debatable. 

Stéphane Smolarek confirms that “the data will go to A/B test”.

The final word:

The acceptance of data, or how to make data more responsible. 
Is it already an issue, to make it socially responsible and to make the field of possibilities a reality?

Romain Charles: “This is a current issue for all of us! What is paradoxical is that according to a McKinsey study, 3/4 of shoppers prefer a brand that will personalize their communications and offers. On the other hand, the shopper wants communications to be personalized but without seeing that they are personalized. Hence the subtlety.
We must therefore be vigilant but not oppose privacy to personalisation. Why not? Because as long as there is consumer acceptance, as long as you do it in the right way, non-intrusively, and as long as there is a legitimate interest for the retailer, it is obvious.
And to go even further in the use of data: tomorrow it can help us to go further on all the themes of responsibility, CSR…”.

Thank you all for your attention, your presence on the day and after thanks to the replay and for your useful feedback.
We hope you enjoyed the format and that the discussions provided you with the necessary and sought-after insights.
See you at the next Lucky Cart talk. Stay tuned!

OUR SOLUTION

More articles