We are introducing Research campaigns in Point11 Analytics. You cannot ask every customer which option they would choose, but you can ask their digital twins. Describe what you want to learn, pick the customers to interview, and Point11 interviews their twins one question at a time, then reads every interview together to show you which option they chose and why.
Research is one of three campaign types in Analytics. Discovery shows how often AI assistants name you. Experience shows what happens when shoppers like yours try to buy from you. Research asks the question behind your next decision.
Built on your digital twins#
Research runs on the same digital twins as the rest of Point11. Point11 builds them from your website, starting with your business profile and a realistic mix of your market, or from your CRM. Each twin is written as a complete person, with a background, needs, constraints, and a reason to buy. When you interview a twin, it answers from that life.
Put the question to your customer models#
The Research launcher asks one thing: "Which one would your customers choose, and why?" You name the study, describe what you want to learn in plain words, and choose who to interview: individual twins, an audience, or, if you have stores, your store customers. You can attach files, such as a concept or an offer sheet. Alpha reads them when it drafts the study, and each twin sees a file once the question that introduces it comes up.
Press Review, and Alpha, Point11's AI digital marketer, drafts the whole interview from your description: the questions, their answer types and options, their order, and where each answer leads. A separate reviewer model checks the draft against your brief before it reaches you. You can reword or reorder any question and add your own, and Fill with Alpha completes any placeholder you have not written yet without touching what you have.
You can also add a task. Give it a start page and success criteria, and each twin's computer-use agent tries it on your website and is judged against those criteria. You see what twins would do, not just what they would say.
Before you start, Preview runs one twin through your questions so you can read a sample interview. Then run the study once, weekly, or monthly. A recurring study gives each interview its own time on your calendar, inside the hours you set, and the same twins answer again each period.
The interviews#
Each interview is a conversation between two models. An interviewer asks one question at a time, written from your guide and everything said so far. The twin answers in its own words.
The design keeps the answers honest. The twin never sees your study goal, the questions still to come, or how its answers will be coded. The interviewer does not read your answer options out as a list; it asks the way a person would. When an answer should change the path, the interview follows the route you set.
Every transcript is there to read. Open any twin to see the whole conversation, question by question, and how its task went.
How findings are summarized#
When the interviews are in, Research reads each one on its own, then reads them all together. Results come back question by question, in the order you asked. Each question shows how many twins answered, a chart that fits the question (a pie for a single choice or a ranking, bars for checkboxes, columns for a scale, completion for a task), a written summary of what the twins said, and quotes. Every quote is an exact passage from a twin's answer, checked against the transcript before it is saved.
Narrow the results to a segment or a location, and the counts are recalculated for that slice. A recurring study compares each wave with the one before it, as long as the questions have not changed, so you can see what moved.
Take the illustrative example on our Research page. An auto insurer asks its customers' twins whether they would switch to a six-month term if the renewal price held. The answers split into would switch, would compare first, and would stay, and the interviews explain the split: a renewal that went up with no claim is what sent customers shopping, and how a carrier handles claims is what would hold them back.
When you are ready to act, take the study to Alpha. Alpha can read the study's interviews and findings and plan the next step with you.
What it is for#
Research is for the decisions you cannot test on every customer first:
- Offers. Which version customers would pick, and what would make them switch.
- Pricing. How customers react to a proposed price, and why.
- Roadmaps. Which features customers want first, and what the list is missing.
- Messaging and positioning. What customers take away, and where it snags.
- Brand. How customers move from knowing your brand to choosing it.
Twins won't predict exactly what every customer will do. They let you explore before you decide. Because the same twins can answer again, you can change an offer, a page, or a message and see what changes.
Start your first study#
Research campaigns are included with Point11 Analytics: a monthly allowance, then per interaction, with each delivered interview counted as one. There is no campaign minimum. See pricing for details.
Create your account, open Campaigns, and choose Research. Describe what you want to learn, pick your customers, and press Review. To see the whole flow first, watch the Research film or book a demo.
