Relationship marketing: what will matter in 2027
Damien Bouquet · CEO Pulp’in
At a recent webinar, two CRM leaders, one in travel and one in consumer goods, shared their current projects and their roadmap. Two very different industries, and yet a lot in common. Here are the takeaways, along with a few recent figures.
50M
RCS-compatible smartphones in France, up from 28.3M a year earlier (af2m)
5–10×
more clicks on RCS than on classic SMS, according to industry players
≈10%
of searches already went through AI by late 2025 (SimilarWeb)
68%
of US consumers often wonder whether what they see is authentic (Gartner)
The welcome booklet is disappearing
In travel, practical information used to arrive in a digital booklet sent before the stay. Few people read it end to end, and questions ended up on the phone.
Today, a conversational agent reaches out to the guest five or six days before arrival. Guests ask their questions when they need to: the parking code, available extras, reception opening hours. The booklet content is still there, built into the agent, which answers from information approved by the brand. That is what makes it reliable.
The most interesting part is how the saved time is used. Teams spend less time on the phone and more time with guests already on site. Here, automation gives teams time back for hospitality.
There is also a clear commercial opportunity: a few days before leaving, guests are picturing their trip. It is a good moment to offer an extra, an activity, or an upgrade, right in the conversation.
What we see
Pre-stay is also where interactive hospitality journeys perform best: guests see their stay, their room, their dates, and pick a late checkout or a dinner on the spot. Every choice reaches the hotel’s tools before they arrive.
Product page bots that explain better
In consumer goods, the issue is understanding. Take a machine sold either outright or with a subscription. Many customers struggle to see the difference, hesitate, and leave without buying. A conversational exchange on the product page clears that kind of blocker in two or three questions, where a comparison table does not always do the job.
Two or three well-asked questions clear a doubt that the best comparison table leaves standing.
RCS is finally taking off
RCS, the successor to SMS, was long a topic for the future. Not anymore. According to af2m, compatible smartphones in France went from 28.3 to 50 million in a year, and brands increased their RCS campaign volume by 75% over the period. The message lands in the usual messaging app, with nothing to install.
28.3M
RCS-compatible smartphones in France a year ago
50M
compatible smartphones today
+75%
RCS campaign volume over the same period
Feedback shared during the webinar points the same way: click rates are very good, and above all the format adds a conversational layer. You can explain a product, a machine, a coffee range. Industry players mention click rates 5 to 10 times higher than classic SMS, consistent with what was discussed.
WhatsApp, for its part, often comes up after the stay or after the purchase, to keep in touch with the customer until the next booking.
Conversational product page
Two or three questions to choose between buying outright and subscribing, right when the doubt appears.
RCS and agent
Practical information on demand, and an extra offered while the guest is picturing the trip.
A relationship kept alive until the next order or the next booking.
One piece of content, every channel
With Pulp’in, the same interactive experience opens from email, SMS, WhatsApp, or RCS, adapted to each channel of the conversation.
Personalizing for each customer
Both speakers used almost the same words: personalization at scale, for each individual. One of them even described his dream: one customer journey per consumer.
The dream: one customer journey per consumer.
Two paths kept coming up:
- → Using LLMs and better customer knowledge to go beyond plain product descriptions, with content that matches what the customer feels or is living through at that moment.
- → Using data that is often already available and underused: the expected date of the next order, churn risk, purchase history. With these signals, you can trigger the right message at the right time and cross-sell or upsell accurately.
| Signal already in the CRM | What it can trigger |
|---|---|
| Expected date of the next order | A reminder when the customer needs it |
| Churn risk | A retention message before they leave |
| Purchase history | A cross-sell or upsell that fits their habits |
Three signals mentioned during the webinar, often available and still underused.
On omnichannel, one figure was mentioned: customers who use several channels would represent about 20% more value. Treat it with caution, as it comes from field experience rather than a study, but it confirms an intuition many teams share.
+20%
value for multichannel customers, according to field experience shared during the webinar
1 = 1
one customer, one journey: the goal both speakers described
Content production, the real bottleneck
Personalizing customer by customer means producing far more content. So the question came naturally: how can CRM teams produce visuals and messages faster? Over the next 18 to 24 months, both roadmaps prioritize tools that save teams time. Briefing an agency for every variant is no longer sustainable.
Our approach
This is the problem Pulp’in chose to tackle: each customer receives their own interactive experience, generated in one second from their data. Variants run into the thousands, production stays that of a single campaign.
Getting cited by AI
Another project is coming fast: visibility in ChatGPT, Gemini, or Perplexity. According to SimilarWeb, about 10% of searches already went through AI by late 2025, and that volume could overtake Google and Bing as early as 2028. This is called GEO, for Generative Engine Optimization.
≈10%
of searches went through AI by late 2025
2028
the year this volume could overtake Google and Bing combined
One of the brands is preparing a new website designed partly for this: better structured, better connected to its data. Also worth keeping in mind: a Kinsta study published in September shows that consumers use AI to shortlist, then visit the brand’s site to check. The website remains the place where you convince.
Keeping the promise of the real
This is where the discussion was richest. In travel, you already feel on holiday when you book. If photos or videos look fabricated, the promise collapses. Customers want to see the real campsite, the real pool.
68%
of US consumers often wonder whether what they see is authentic
1 in 2
prefers brands that do not use generative AI in their communication
Studies point the same way. According to Gartner, 68% of US consumers often wonder whether what they see is authentic, and half say they prefer brands that do not use generative AI in their communication. In Europe, the AI Act now requires AI-generated content to be labeled.
The speakers took a nuanced position: use AI where it saves time and adds relevance, keep what the customer sees real, and always leave the option to talk to a person.
The real, assembled for each customer
Pulp’in assembles the brand’s real photos and videos with each customer’s data. The campsite, the pool, the room the guest will actually see: automation handles the assembly, the images stay authentic.
Key takeaways to prepare for 2027
- → A conversational agent fed with your own content can advantageously replace the booklet or the FAQ, as long as the saved time goes back into hospitality.
- → Conversations before purchase or arrival are a good moment for upselling.
- → RCS deserves a serious test, used to converse and explain rather than to push promotions.
- → WhatsApp and RCS help keep in touch after the purchase.
- → Your CRM data probably holds untapped signals (next order, churn).
- → Without a fast production tool, individual personalization will remain wishful thinking.
- → Your site must be readable by AI and convincing for visitors who come to check.
- → Transparency about AI, real visuals, and a human contact available: that is what preserves trust.
Sources (in French)
- → RCS passes 50 million compatible phones (Clubic, af2m data)
- → Mobile marketing in France (Klaviyo)
- → Half of US consumers prefer to avoid brands using generative AI (Next, Gartner study)
- → Faced with AI, consumers favor human recommendations (The Media Leader)
- → Being visible in AI chatbot search (Hospitality ON)
- → Kinsta report on AI use for search (September 2026)
Frequently asked questions
Yes. According to af2m, 50 million smartphones are compatible in France, up from 28.3 million a year earlier, and RCS campaign volume grew 75% over the period. The format works best when used to converse and explain.
GEO, for Generative Engine Optimization, covers the practices that make a brand visible and cited in answers from ChatGPT, Gemini, or Perplexity. About 10% of searches already went through AI by late 2025.
By designing a template once, then assembling it automatically with each customer’s data. That is the principle behind Pulp’in: one interactive experience per customer, generated in one second, from the brand’s real visuals.