For entrepreneurs to succeed with their buyer advertising efforts, it’s important to grasp which prospects are blissful, that are liable to churn and which current cross-sell and upsell alternatives.
Your buyer knowledge is stuffed with clues that can assist you perceive which prospects match into every of those buckets. You simply have to mine that knowledge to seek out the clues. Sounds easy, proper? It hasn’t all the time been that approach. Then alongside got here generative AI.
If you happen to’re like most individuals, you hear the time period “genAI” and you consider content material creation. However its capabilities are increasing and customers are experimenting with extra use circumstances.
For this text, I used the Google Gemini generative AI utility to assist me analyze buyer knowledge, establish prospects that met sure standards and obtain suggestions and messaging to make use of with these prospects.
Creating pattern buyer knowledge with Google Gemini
Earlier than I may analyze buyer knowledge, I wanted to create it. I constructed a state of affairs with Gemini to primarily fabricate knowledge in a spreadsheet for patrons of a small B2B software program firm
I informed Gemini the columns I needed within the spreadsheet and supplied ranges for among the knowledge in an effort to stop some prospects having lifetime income a number of a number of greater than others, for instance.
The columns I selected to create have been:
- Firm title.
- Variety of licenses for 2003 and 2004.
- Complete time logged within the app (2003 and 2004).
- Common time logged within the app (2003 and 2004).
- Lifetime worth.
- Buyer helps prices (2023 and 2024).
- Common month-to-month assist prices (2023 and 2024).
One of many memorable moments from this a part of the train was once I determined so as to add year-over-year knowledge so I may run comparisons. Gemini added 2023 and 2024 lifetime values for every buyer. I identified that there’s just one lifetime worth, and Gemini instantly apologized for the error and glued it.
All informed, Gemini created pattern knowledge for 150 prospects. Right here is the pattern knowledge for eight of them.
Let’s dig into buyer assist prices
You possibly can be taught quite a bit by analyzing assist prices. Prospects with excessive assist prices are in danger for churn. And prospects that value a vendor extra in assist than they create in in income are dangerous offers that should be fastened.
That’s why my first question was to ask Gemini how lots of the 150 prospects had greater assist prices than income. Happily, none of them did, which is nice information. Subsequent, I needed to know which prospects had the best assist prices over the 2 years of knowledge we created.
Observe that for annually the identical firms had the best assist prices, which might be an anomaly in our buyer knowledge, a misunderstood immediate or it might be an indication of a bigger downside with the shoppers in query.
What do I say to prospects whose assist prices are killing us?
Since Gemini is a generative AI platform, it’s greater than able to taking the info and evaluation and drafting related communications. So I requested Gemini to draft an electronic mail to ship to prospects with excessive assist prices providing them assets to assist.
Gemini’s preliminary draft cited the excessive assist prices within the electronic mail, which I requested it to right as a result of it’s not smart to inform your prospects “You’re costing us a ton in buyer assist!”
The second draft was nearer to what I used to be searching for:
Expensive [Customer Name],
We needed to achieve out to specific our appreciation in your continued partnership with [Company Name]. We worth your online business and are dedicated to offering you with the very best assist.
That will help you get probably the most out of our services and products, we’d prefer to give you entry to further coaching and assets. These might embrace:
[Training Course 1]: [Brief description of the course]
[Training Course 2]: [Brief description of the course]
[Knowledge Base Article]: [Link to a relevant article]
These assets may also help you and your group troubleshoot issues extra effectively, prepare new customers successfully, and maximize the worth you get from our merchandise.
By profiting from these assets, you possibly can doubtlessly cut back the necessity for assist interactions and optimize your total assist expertise.
Please tell us in case you’re taken with studying extra about these assets or in case you have every other questions. We’re right here to assist your success.
Let’s discover these churn dangers and get ’em fastened
I requested Gemini to establish within the knowledge the 5 prospects most in danger for buyer churn. It identified that it was able to hypothesizing which prospects have been most in danger, however that further knowledge, like historic churn charges for instance, would assist establish dangers extra precisely.
Right here’s a pattern from the 5 prospects it recognized as churn dangers and why.
Figuring out churn dangers is barely half (and even much less) of the battle. So I requested Gemini to assist me define a retention technique for these prospects.
A loaded query: Who’re the ‘greatest prospects’?
Everybody in buyer advertising would like to know which of their prospects are “the perfect prospects.” However quite a bit goes into defining a “greatest buyer.” And, as I anticipated, once I put the query of which prospects from our dataset have been “the perfect” and why, Gemini jogged my memory we’re working with a comparatively easy dataset.
It may do a fair higher job of answering the question, it stated, if it had data on:
- Buyer satisfaction rankings.
- Product utilization patterns.
- Churn historical past.
- Referrals made.
However, Gemini took a shot at figuring out the perfect prospects based mostly on the info we had round buyer lifetime worth (CLTV), assist prices and engagement with the product.
What I discovered from this train
The pure language capabilities of Gemini and different genAI purposes are getting higher. I didn’t have to create sophisticated prompts or ask Gemini to play a job. I merely requested it to do what I needed it to do.
Greater than spitting out solutions, nevertheless, Gemini added helpful strategies, comparable to further knowledge for our dataset that may be useful, or strategies round methods.
I discovered Gemini’s position on this train to be half simulator and half mentor. We have been utilizing fabricated knowledge, and whereas the train was fictitious, it was additionally very actual. This might have been precise buyer knowledge and the outcomes and strategies would possible maintain up. Whilst a simulation, it made for an excellent thought train.
The strategies and areas for enchancment Gemini provided have been much like working with a extra skilled mentor. Gemini was proper in lots of circumstances. I didn’t add knowledge like buyer satisfaction scores, for instance, or referrals. Nor did I take a shot at including buyer acquisition prices. That’s the kind of suggestions a extra skilled marketer would possibly ship in a case like this.
My present plan is to maintain the info on the 150 fictional prospects and add to it. I’ll proceed to ask Gemini to provide me insights and strategies. I can’t wait to see what I be taught alongside the best way.
Dig deeper: Meet my analysis group: Gemini, ChatGPT and Perplexity
