Your emails ask questions. And get answers

Your emails ask questions. And get answers

As of November 2026, the shortest answer to "can an email collect data instead of asking for a click?" is yes, on a narrow list of mail clients. Optimizely published a feature page on 10 November 2026 describing interactive emails: a recipient rates, votes or fills a field inside the message, and the answer lands in its Campaign product, ready for segments and automations. Anyone weighing an ai automation for business programme for 2027 should read that claim carefully. The collection is what changes; the automation underneath works as it always did.

Key takeaways

  • Interactive elements work in Apple Mail on iOS and macOS, Samsung Email and other modern clients; every other client shows a fallback that opens a landing page.
  • Optimizely sells the elements as a package, with a stated go-live window of 14-30 days and an IT effort of 0 hours on the customer side.
  • Its own case study reports one mailing collecting 41.3x more data in under a day, at a 4.9% completion rate among profiles that were previously incomplete.
  • The case is a vendor case, not independent research, so treat the multipliers as a ceiling, not a benchmark.
  • The harder question is not collection but ownership: where the answers live, and who can segment on them.

What exactly happens when someone clicks a star inside an email?

The mechanics are worth spelling out, because most marketing teams picture a landing page with a form on it. Optimizely's page describes a different path. On a modern client the star, the scale or the field renders inside the message. The recipient taps it, and the value is written into Campaign without an export step and without an integration to build.

On Gmail, Outlook and the rest, the same mailing shows a fallback. Where the point of the message is collecting data, a button opens a landing page that Campaign builds, and the answer still ends in Campaign. Recipients on those clients take one extra step. That is a real difference in completion, and the vendor's headline numbers come from a mix of both paths, so a Singapore list dominated by Gmail and Outlook will not replicate them exactly. The same question comes up whenever teams compare site forms with conversation channels, and the 2027 local search article on this site treats the channel switch as a design constraint rather than a detail.

The package carries five interactive elements, which the vendor says can be reused in identical form across mailings: a star rating with a follow-up question for low scores, an NPS element on a 0 to 10 scale, a form and survey block with multiple fields, an image slider holding up to 5 linked images, and a collapse tool for hiding text and images. A repeated NPS question overwrites the earlier answer, which keeps the record current rather than appending duplicates. For a team with no engineering time, that reuse is the whole argument: one set of questions can run in every newsletter, and the answers keep landing in the same place.

Did the numbers come from Optimizely or from an independent study?

They come from Optimizely, and the page says so plainly: this is the Dorothee Schumacher case study, a fashion brand using the vendor's own product. Quote it as what the seller reports, not as market evidence.

The starting position the vendor describes: the birthday automation was one of the strongest ROI drivers in that brand's CRM, running at an open rate of around 50% and a click rate of roughly half that, and yet 68.7% of the full mailing list had no birth date on file. Organically, about 4 new profiles arrived each day. At that pace the gap would have taken years to close, and classic web forms cost clicks, cause drop-offs and tie up IT capacity. That is a familiar situation for any Singapore SME running ai automation for business on a list it has owned for years: the automation is fine, the record underneath it is thin.

One interactive mailing changed the arithmetic. In less than a day it collected 41.3x more data than in the preceding period, which the vendor also expresses as 240x faster data collection per day. The completion rate among previously incomplete profiles was 4.9%. Data coverage moved by +3.3 pts.

Run those last two numbers past a colleague before you budget on them. A 4.9% completion rate among people who had never given the date is a modest conversion, and the coverage gain is 3.3 points of a list that was mostly empty. The 41.3x multiplier is large because the denominator was almost zero. If your list already has most of the fields you need, the same campaign has far less room to impress.

What is really inside the 14-30 day window?

The package ships with a stated go-live window of 14-30 days, an IT effort of 0 hours on the customer side, and simple changes allowed up to 4 times a year. The page also notes first-level consulting from Optimizely. Read together, those lines describe a service engagement with a delivery team behind it, not a toggle you switch on.

For an SME this is the part to interrogate in the sales call. A window of 14-30 days is a range, and the page does not say what determines where in it you land. The 4-changes-a-year limit sets how much the survey copy can move once the package is running; a business that rewrites questions monthly will find that ceiling before long. The 0 hours figure is the vendor's statement about the Dorothee Schumacher use case, and it holds only while you accept the elements, the mailing and the storage as the vendor builds them. Change the flow, connect your own warehouse, or need the answers in a second system, and the number stops describing your project.

None of that is a criticism of the product. It is a description of what a package is: a fixed set of parts, delivered on someone else's process. Whether that is the right shape depends on how much of the data path you intend to own, and that question belongs in the same conversation as any other ai automation for business decision.

Should you collect first-party answers in the inbox or on your own site?

This is the decision the news really poses, and it splits cleanly by what you already run.

Collect in the inbox when your list is the asset, the answers are simple (a date, a score, a preference), and the automation that consumes them already lives in the same marketing platform. The economics are strong: no click, no redirect, no lost session, and the answer is usable the same day.

Collect on your own site or through your own data layer when the answers must feed several systems, when you need the record in a warehouse you control, or when the questions change often. A web form you build costs developer time and loses people at the click, which is exactly the cost Optimizely is selling against, but it keeps the data path in your hands.

A practical test: ask who will be able to segment on the answer in six months. If the answer is "the marketing team, inside one tool", the in-email route fits. If it is "whoever queries the customer database", the collection point is less important than the destination. That is where web application development earns its cost, because a connector or an API you own keeps the answer in a record several teams can read.

Which collection points feed your automation today?

Most SMEs already answer this question, they just have not written it down. Their answer data arrives from a handful of places, and the list usually looks like this.

Collection pointWhat arrivesTypical owner
Website form or landing pageEnquiries, preferences, survey answersMarketing, with developer time
Chat window on the siteIntent, contact details, questionsSales or support
Native ad lead formsName, contact, sometimes a qualifierPaid media team
CRM recordsStage changes, call notes, deal dataSales operations
Email bodyRatings, NPS, short answersNobody, until a feature like this arrives

The last row is the new one. Everything above it is mature, documented and usually connected. The news adds a row at the top of the funnel where attention already sits, and it does so without a developer, which is why it gets attention from teams with no engineering capacity.

If you are planning the automation layer rather than the collection layer, note the distinction the CUHK School of Continuing and Professional Studies course page draws in its own description: practical business automation works as trigger, decision and action, which is the same skeleton whether the trigger is a form submission or a tap on a star in an email. The JieGou glossary makes a parallel split between rule-based automation that moves data between apps and AI-driven automation that handles ambiguity. In-email answers are the first kind of data; what you build on top decides whether you get the second kind of behaviour. For an SME in Singapore, that usually means one small custom piece of plumbing rather than a platform replacement.

What does Topkee do about this, and where does it stop?

Topkee's position on this news is specific, and it is worth being blunt about it: the knowledge base holds no email-marketing or interactive-email product. Its collection points are website forms, chat, native ad lead forms and the CRM, not the email body. A team that wants ratings and NPS collected inside a mailing has to buy that from an email platform, and Topkee is not one.

What Topkee builds is the layer underneath, the one that has to receive those answers once they exist. TTO CDP is described as the customer data platform and marketing attribution product used for customer management, platform data connection and data standardisation, with three stated capabilities: connect platform data, clean up toxic data, and standardise data standards. Its integrations catalogue already lists the routes answer-type data takes, including HTTP API, General Website Form, Leadbox Webhook, and lead-form webhooks for Facebook, Google, TikTok and LinkedIn. That matters here because a rating, a vote or a filled field is event data like any other data point, and the catalogue is where such an event can enter a single governed record rather than a silo inside whichever tool captured it.

On the collection side, Leadbox covers forms and leads: sales teams can customise sales stages and event tags without development assistance, and when a lead's stage changes the system triggers the event and passes it back through TTO Tag for attribution. For the question of where to collect, the honest summary is that Topkee's chain runs from data collection through unified governance to audience segmentation and automated follow-up, but its collection sits in web forms and conversation channels rather than in the inbox. A company that expects one vendor to cover both ends will be disappointed; a company willing to join two vendors at the CDP will not be.

Do you need a custom application to automate on the answers?

Sometimes. The rule of thumb we use: buy the collection, own the logic that is specific to you.

If your automation is a birthday message, a follow-up on a low score, or a segment for the next newsletter, a packaged feature and a platform automation will carry it. Building an application for that is web application development spent on work the market already sells.

If the automation has to read an answer, decide something about the customer, and act across systems that do not share a data model, the packaged route runs out. That is where the difference between rule-based tools and AI-driven workflows starts to bite. A rule reads a field and sends a template. A system that reasons over the answer can route a low score to a different team than a high one, and can do it without a rule for every combination.

A useful way to frame the choice, and the one worth bringing into a 2027 budget discussion: cost the collection and the logic separately. Collection is increasingly a feature you buy, and the price is falling. Logic is where your business differs from the next one, and it is the part worth owning. In practice that means a Singapore SME can budget the email feature as a line item on a platform invoice and the logic as web application development work scoped to its own data model, instead of asking one supplier to quote both.

What should an SME do first?

Start with an audit of the fields you are missing, not with the channel. If a single campaign would only move coverage by a few points, the interactive route is a small win and should be treated as one.

Then check your client mix. If most of your list reads mail in Gmail and Outlook, a large share of your recipients will take the landing-page path, and the completion numbers will look closer to a web form than to the vendor's case.

After that, decide where the answers must live, and make sure the collection tool and the data platform agree on that destination before anyone signs. The AI advertising readiness audit on this site is a reasonable starting point for a team that has pilots running and cannot say what data they produce.

The last step is the least glamorous and the one most teams skip: write down who owns the record. An email platform that stores the answers is not the same as a customer database that governs them, and the difference only shows up when someone asks for a segment next quarter. This is where an honest ai automation for business plan separates the parts you buy from the parts you keep.

Why does the market answer "ai automation for business" with Unisoft ai 課程?

Search the phrase in Singapore and Hong Kong today and the result page is largely training. Queries in the same space, such as Unisoft ai 課程, sit in a result set full of course listings and workshops alongside vendor glossaries. For a buyer with a budget, that is a confusing signal: the dominant offer is instruction, not implementation.

The CUHK School of Continuing and Professional Studies runs a part-time course on smart business automation with AI, built for owners, operations managers and customer service supervisors rather than developers, and it uses a no-code visual tool to assemble workflows. Its own description sets out a framework of pain point, process, people, tools and cost, and the course ends with participants drafting an ROI analysis for their own organisation. That is a sensible way to learn what to automate. It does not deliver a working system, and it does not collect data on your behalf.

The JieGou glossary sits in the same part of the results: definitions, comparisons between rule-based and AI-driven automation, and framing by department. Useful background, no implementation.

So the gap between the search results and the buyer's problem is this. The market will teach you the vocabulary of automation and sell you a course on it. Fewer pages explain what to buy first, and which parts of the data path you should keep in-house. A marketing team that finishes a course still has to decide whether to collect first-party answers by email versus the site, and who will own the record afterwards. Those are procurement questions, not curriculum questions, and they are the ones this article sets out.

FAQ

Can an email really collect an answer without a click to a landing page?

Yes, on Apple Mail for iOS and macOS, Samsung Email and other modern clients. On Gmail, Outlook and the rest, the mailing shows a fallback where a button opens a landing page, and the answer is stored the same way.

Is this available to businesses outside Optimizely Campaign?

No. Optimizely sells the elements as a package for Campaign customers, with a go-live window of 14-30 days and first-level consulting attached to the package.

What kinds of question can an interactive email ask?

Optimizely's package covers star and emoji ratings, an NPS score on a 0 to 10 scale, multi-field forms and surveys, an image slider with up to 5 linked images, and a collapse tool. Free-text answers are not described on the page, so a customer comment box is not part of the offer.

Is the 41.3x figure something I can expect?

Treat it as a vendor case result, not a forecast. It came from one fashion brand whose list was 68.7% missing a single field, where organic collection was about 4 profiles a day. The completion rate among those profiles was 4.9%, which travels better to a normal list than the multiplier does.

Who should not adopt this yet?

Teams whose answers must feed several systems, whose questions change more than 4 times a year, or whose audiences open mail mostly in clients that show the fallback. For them, the collection point is solving a problem they do not have.

Sources

  1. "Your emails ask questions. And get answers", Optimizely, 2026-11-10. https://www.optimizely.com/products/campaign/feature-update-interactive-email
  2. "Smart Business Automation with AI (AI智能業務自動化實戰)", CUHK School of Continuing and Professional Studies, undated. https://www.scs.cuhk.edu.hk/en/part-time/it-fundamentals-applications/smart-business-automation-with-ai-ai/262-191320-01
  3. "What Is AI Automation? — Using AI to Execute Business Processes", JieGou, undated. https://jiegou.ai/zh-TW/glossary/ai-automation/
Category:SEO & AI Search
Share to:
Date: 2026-11-12
Wing Yau

Article Author

Wing Yau

Marketing Manager