Case study · AI hyper-personalisation
“A truly personal letterfrom your favourite artist.”
How Meister Lampe und Freunde used AI-driven hyper-personalisation to significantly increase sales of limited sketches for artist Lars Stenchly – turning followers into engaged supporters.
Situation & goal
Almost 1 million followers – and a suspended account.
Lars has a combined reach of almost 1 million on TikTok and Instagram. When his TikTok account was temporarily suspended in 2023, he started building an email list with the help of Meister Lampe und Freunde.
That list now holds 5,000 addresses – a fraction of his social reach. Yet it sells more works than all social channels combined. Email conversion is roughly 150 times higher than on social media, despite his young audience.
The goal: sell limited sketches to visitors of his online gallery – originals positioned as a “middle option” between affordable art prints and the expensive original works.
The lever: every customer is addressed individually, based on their interests. No generic advertising, but a message showing that the artist truly understands his fans and their preferences.
Results
How to multiply your click rate by ten.
Recipients were split into three segments receiving different levels of personalisation. The hyper-personalised segment outperforms the others on every metric by a wide margin.
Open rate
Click rate
Click-through rate
Open rate 2.1×, click rate 10.4×, click-through rate 5.0× higher than in the non-individualised segment.
Approach
From survey to engine.
Data collection
A targeted survey among Lars’ subscribers captured emotional preferences: preferred art styles, the importance of size and format, the relevance of the story behind a work, original versus limited edition. As a thank-you, participants could win a signed art print.
Data preparation
Survey results were combined with purchase behaviour, click history and demographics – a comprehensive picture of each art lover, the same way any online shop can use it for targeted recommendations.
MLHPE
On this data basis, the Meister Lampe Hyper Personalization Engine was built using the ChatGPT API – an AI engine that creates tailored content at scale. Intensive prompt engineering included.
Past purchases, preferred products, repeat orders – the basis for fitting recommendations.
How often a contact interacts with content and which links interest them – measured in this project as “LARS”: the Lars Audience Reactivity Score.
Preferences and values captured by survey – in an artistic context the key to a deeper bond.
Language, regional and gender-specific preferences shape how content is structured and phrased.
The engine
Meister Lampe Hyper Personalization Engine.
Fully automated personalised content that responds to individual customer preferences.
The MLHPE combines customer interactions, demographic information and the survey data to create unique copy tailored to each recipient – from subject line to P.S. A lover of classical art reads something different from someone fascinated by modern forms.
This case study focused on copy. But the engine is also the technical foundation for individualisation in other channels – making it broadly applicable in e-commerce and online marketing.
Result
Unique copy – for every recipient.
Three examples from the campaign. Each email addresses the recipient by name and mirrors exactly what they said in the survey.
Hi Johanna,
as a lover of classical art, I have a few sketches that will surely match your taste. Each piece is unique and brings a fresh perspective to traditional art. Thank you for appreciating my work – I hope these pieces speak to you too.
Hi Isabella,
I’m delighted you follow my emails so actively! Since you value the stories behind art, I have a few small sketches that stir big emotions. Each piece is an original that tells its own story and is waiting to be discovered by you.
Hi Christopher,
thank you for your last purchase! As a connoisseur of classical art and original works, I’ve created a few sketches that will surely match your taste. Each piece is a true original capturing the depth and elegance of classical art.
GDPR
Hyper-personalisation, privacy-friendly.
The MLHPE shows that hyper-personalisation can be implemented in a GDPR-compliant way: personal data such as email addresses are replaced by anonymised IDs. Only anonymous ID numbers and datasets reflecting interactions and preferences are processed.
Because the original personal data is never linked to the content of the personalised messages, the engine uses only the data required to create the content – without any direct identifiers.
Hyper-personalisation for your business?
The approach works wherever first-party data meets individual customer needs – in e-commerce and B2B alike.
Die Highlights