Max.AI is your new, custom-built chat assistant. A Dr.Max email specialist asks it an ordinary question (e.g. “Which campaigns had the highest open rates last year?”) and within moments gets an answer computed from the real data of every email campaign ever sent, along with a recommendation on what to do differently next time to get even better results.
Before every send, the email specialist goes through the complete campaign history:
They then build the new campaign on what did (and didn’t) work, ensuring continuous growth of the email channel.
In reality, the specialist has neither the time nor the analytical backing for such advanced analysis, so they decide by intuition.
But that breaks the most important emailing best practice and ignores the most valuable asset: the rich data collected over years of sending.
That is why we built Max.AI for Dr.Max: a custom application that solves this problem and, on top of that, gives the specialist feedback on all of their work.
Where it’s already live and what’s next ↓
Salesforce Marketing Cloud logs every send: opens, clicks, unsubscribes and A/B test results. But the numbers are scattered across hundreds of sends and reports that have to be clicked through, exported and stitched together one by one. On top of that, someone has to match the numbers with each send’s content and segment, i.e. what was actually behind the results. Pulling together, say, every dermo send from the past few years means hours of work. Max.AI has it all in one place: data, content and segments, and interprets them within seconds.
When next Wednesday’s newsletter is being prepared, hardly anyone has time to check how the same topic performed across the previous 18 sends, which subject line won back then and which recipient list to use. The experience lives in people’s heads, not in the data. Max.AI pulls up the entire history of a topic with a single question and, based on past results, directly recommends the subject line, the send day and the recipient list.
We built an assistant on top of your own data that speaks the language of your market: Czech, Slovak, English, Serbian, Polish, Romanian or Italian. It looks up the data in the database itself, computes the result and cites the numbers it actually found. The specialist talks to it like to a colleague and gets an answer backed by data. On top of that, we develop Max.AI according to our years of emailing know-how, so it knows what a perfect specialist knows. And whatever we cover with your people in workshops, we immediately teach Max.AI too.
It is a native application developed specifically for Dr.Max. At the same time it works like the classic chatbot users already know, so it feels friendly and understandable from the first minute. Conversation history is saved and can be revisited at any time.
The specialist asks in plain language, in the language of any market: “Which day of the week do our emails get opened the most?” No formulas, no filters, no knowledge of the data structure needed. Max.AI answers that most opens come on Wednesday, with an open rate of 35.3%.
In the background it builds and runs a query against the campaign database, compares variants and computes averages. It does the work a data analyst would otherwise do, and unlike ordinary AI chatbots it does not make results up: every number in the answer comes straight from the database.
The answer arrives as a clear table with a benchmark comparison and concrete advice for the next campaign. And Max.AI does not stop at the question asked: in its recommendations it also points out connections the specialist would not have asked about, because they simply would not have occurred to them.
Send data from Salesforce Marketing Cloud is loaded into Max.AI’s own database, where it is enriched with campaign topics, A/B test results and the team’s notes. On top of the database runs the Claude language model by Anthropic, the same technology behind the most demanding enterprise AI applications. The model translates the user’s question into a database query, runs it and builds the answer exclusively from the numbers the database returned.
And above all: it is enterprise ready. The data stays in an isolated instance operated for Dr.Max by Data Club and never leaves the EU. Everything also runs in zero data retention mode: neither queries nor data are stored anywhere and no AI model learns from them. That is a guarantee you will not get with an ordinary AI chatbot or with competing tools. Nor will you get control if everyone on the team uses AI their own way and sends company data to tools nobody has vetted. Max.AI gives the team one safe, vetted place to work with data.
The numbers are real, taken from Dr.Max data for the Slovak market (hence the Slovak campaign names). Open rate = % of recipients who opened the email; click rate = % who clicked a link in it.
Best-performing email subject lines (real sends, 10k+ emails delivered):
| Subject | Topic | Open rate |
|---|---|---|
| Zľavové kupóny sú späť! | Kupónová akcia | 67.6% |
| Náš recept na Valentín 💞 | Valentín | 67.5% |
| Vaše obľúbené dermo značky v akcii | Dermo | 61.8% |
| Cenová bomba Lavazza | Mix akcií 3 | 61.5% |
| Top ponuky pre najmenších | Matka a dieťa | 59.5% |
| Letný výpredaj až do -50 % | Výpredaj | 50.8% |
| Novinka pre členov Dr.Max CLUB | 2+1 zadarmo | 49.4% |
| Vyšetrenie rizika cukrovky v Dr.Max | Deň diabetu | 48.7% |
| Exkluzívne akcie len pre členov Dr.Max CLUB | Top akcie | 48.2% |
| Darček k nákupu Simpl Therapy | Dermokozmetika | 46.8% |
General patterns across the successful subject lines:
Variant B won, and by a wide margin. On the main “LP” promo list (92k addresses) the variants performed like this:
| Variant | Subject | Open rate |
|---|---|---|
| A | Milgamma N, Strepsils či Otrivin za výhodné ceny | 29.6% |
| B | Výhodné ceny na Milgamma N, Strepsils či Otrivin | 38.5% |
A mere change of word order, moving “výhodné ceny” (great prices) to the front, brought +8.9 percentage points, roughly 8,000 extra opens. I recommend saving the “benefit first” pattern as a note on this send so it is not forgotten next time.
The real power: the manager learns this in 20 seconds, not after an hour in an export.There are 6 sends planned. The strongest topic is Wednesday’s Letný výpredaj (summer sale): sale campaigns have 18 historical sends with an average open rate of 35.9% and a click rate of 3.0%. Last August’s subject “Letný výpredaj až do -50 %” reached a 50.8% open rate on 154k emails. I recommend keeping the “Letný výpredaj + specific discount” format and sending to the main (LP) list.
Friday’s Ženské zdravie (women’s health), on the other hand, has only 2 historical sends with a 28.5% open rate; there I recommend an A/B test of the subject line so we gather data.
The plan and the history are connected: the assistant knows what is coming and immediately attaches what worked.Until now, every campaign started from scratch and the experience stayed in people’s heads. Max.AI turns that around: every send, every A/B test and every note becomes part of a memory the next campaign is planned from. Emailing stops relying on individual intuition and starts to improve systematically.
The specialist also gains a partner who has the complete channel history at hand at all times, answers within seconds and gives feedback on every campaign. Decisions that used to be made by gut feel are suddenly made on data. And because we continuously teach Max.AI our emailing know-how, it gets smarter with every workshop and every campaign.
Slovakia has proven that Max.AI works. The pilot has turned into one clear next step: