
Microsoft 365's Homepage Test: A Small Eyebrow Label With Big Hierarchy Implications
Microsoft 365 is A/B testing a 'NEWS' eyebrow and heading above its promo card row. Here's what the added context does to hierarchy and scan pattern.
Aiko Tanaka
Design Director · · Updated
TL;DR
- Detection: Microsoft 365 homepage (microsoft.com/microsoft-365), surface tested = layout, live A/B test.
- Variant A (visits 2 and 6, both Chrome): adds a "NEWS" eyebrow label plus heading "Discover what's happening with Microsoft 365" above the promo card row.
- Variant B (visits 1, 3, 4, 5, 7, including a third Chrome visit): omits the heading entirely, cards run with no section label.
- Detection method: vendor_sdk_active, analyst confidence 75%.
- Same browser engine produced both variants, confirming a server-side experiment, not a rendering fluke.
- Hypothesis logged: Microsoft is testing whether a contextual header improves engagement and click-through on the promo cards beneath the hero.
What exactly did Microsoft change on the homepage?
Microsoft added a small eyebrow label reading "NEWS" and a heading, "Discover what's happening with Microsoft 365," directly above the row of promotional cards below the hero. In the control, that same card row appears with zero framing — no label, no heading, just the cards. Two of seven captured visits (2 and 6) show the variant; five (1, 3, 4, 5, 7) show the control. Both browsers involved were Chrome, which is the detail that matters most here — it rules out a rendering artifact and confirms this is a genuine server-side split test.
Why does an eyebrow label change how the page reads?
An eyebrow label works as a Gestalt "common region" cue, visually grouping the cards beneath it into a single labeled unit instead of a loose row of tiles. Without it, the promo cards read as decoration trailing off the hero — visually present but conceptually untethered. Add "NEWS" plus a heading, and the same cards get reclassified in the visitor's head as content worth reading, not ads to skip. That's a meaningful shift in perceived value for zero additional pixels of card real estate.
This is the kind of lever Contentful's CRO guide points to when it says layout changes — not just copy or button color — can materially shift engagement, because layout governs how visitors group and prioritize what they see before they read a single word.
Where does the eye land first, with and without the header?
Without the header, the eye lands on whichever card has the strongest image contrast, since there's no textual anchor to intercept the scan first. With the header in place, "NEWS" — short, capitalized, high contrast against body copy — becomes the first fixation point, and the heading beneath it acts as a second, larger anchor before the eye drops into the cards. That's a textbook F-pattern setup: a strong top-left (or top-center, depending on breakpoint) anchor followed by a horizontal sweep across the row. The control skips the anchor and forces the cards to compete with each other for that first fixation, which usually means the leftmost or most visually loud card wins by default rather than by editorial intent.
The label also adds a beat of hierarchy before the cards visually justify themselves — visitors get a reason to look ("this is news, it's current") before they evaluate whether the content is relevant. That reduces the cognitive tax of parsing an unlabeled tile row, especially for returning visitors who may not expect promotional content this far down a homepage they've seen before.
What does the variant breakdown look like across visits?
Here's how the seven captured visits split between control and variant, all under the same detection pass:
| Visit | Browser | Header present? |
|---|---|---|
| 1 | — | No |
| 2 | Chrome | Yes ("NEWS" + heading) |
| 3 | — | No |
| 4 | — | No |
| 5 | — | No |
| 6 | Chrome | Yes ("NEWS" + heading) |
| 7 | Chrome | No |
Two Chrome visits (2 and 6) show the header, one Chrome visit (7) shows the control — same rendering engine, different experience. That pattern is the strongest available signal for a live, server-assigned experiment rather than a caching or client quirk.
Is this consistent with how Microsoft tests elsewhere?
Yes — Microsoft has a long institutional habit of testing small-surface changes at scale rather than shipping by instinct. The most-cited internal example is Bing's ad title test, where a low-priority tweak to headline wording lifted revenue by roughly 12%, worth more than $100M annually, with no measurable hit to user experience, as documented in this case study on brand-scale A/B testing. That result reportedly reshaped how seriously Microsoft treats experimentation across Office, Xbox, and other product lines — testing culture, not just testing tools.
A eyebrow label and heading is a much smaller bet than ad copy on a search results page, but the instinct is the same: treat hierarchy and framing as testable variables, not fixed design decisions. Optimizely's definition of A/B testing makes the same point — the range of what's testable spans from a single headline to a full redesign, and the goal is always shifting the conversation from "we think this helps engagement" to "we know it does."
Common questions
Why would Microsoft test something as small as a heading label?
Small surface changes are cheap to ship and easy to isolate statistically. A single heading addition changes one variable, so any lift or drop in click-through on the promo cards can be attributed cleanly to that change, unlike a full redesign where multiple variables move at once.
How do we know this wasn't just a caching bug?
Both the variant and control appeared in Chrome sessions — visits 2 and 6 showed the heading, visit 7 (also Chrome) did not. Identical rendering engines producing different output on the same page rules out a browser-specific rendering issue and points to server-side experiment assignment instead.
What metric is Microsoft likely optimizing for here?
Given the placement directly above promotional cards, the most likely target metric is click-through rate into those cards, with downstream engagement or feature-page visits as a secondary signal. The eyebrow label functions as a scent-of-information cue rather than a conversion CTA itself.
Should smaller SaaS teams copy this exact pattern?
The underlying principle — labeling grouped content to raise its perceived value — transfers well regardless of company size. The specific wording ("NEWS") is less important than the act of giving unlabeled content sections a textual anchor before asking visitors to scan them.
The takeaway for your next sprint
Before you ship a new promo row, resource block, or card grid this sprint, audit whether it has a hierarchy anchor at all. If your eye has to do the grouping work that a two-word eyebrow label could do instantly, you're leaving engagement on the table — test a label, a heading, or even just an icon-plus-microcopy pairing above the block, and measure click-through the way Microsoft is doing right now.
Read next
More from the ABWatcher blog
Nov 22, 2026
Snake River Farms Tests a $25-Off Popup on Its Ad Landing Page
We caught Snake River Farms split-testing an email-capture form and $25-off modal on its landing page. Here's the math on what it would take to call a winner.
Nov 21, 2026
Snake River Farms Tests Mid-Page Email Capture on Its Ad Landing Page
SRF is A/B testing a mid-page lead-capture block and a $25-off popup against a simpler hero-only promo banner on its paid landing page, detected 2024 week-over-week.
Nov 20, 2026
Attio Is Testing a New Tier, Higher Prices, and a New Pitch
Attio is running a live pricing page test with a new $1,000 "Ultra" tier, higher Plus/Pro prices, and a new headline — here's what we caught and what it means.
Nov 19, 2026
Found Quietly Tested a 25% Price Cut on Its Pricing Page
We caught Found running a live pricing A/B test with an annual billing toggle that drops Found Plus from $35 to $26.25/mo. Here's what changed and why it matters.
Nov 8, 2026
Astro Proves "Zero Lock-In" With a Live Code Demo, Not a Claim
Astro's homepage replaces a trust badge row with a working, clickable framework demo — here's the hierarchy and interaction design behind it, detected 60% rare on astro.build.
Nov 7, 2026
Basecamp's Big Numbers: Why Hyper-Precise Stats Beat Vague Social Proof
Basecamp's homepage swaps "trusted by thousands" for 1.7 billion to-dos and 5 petabytes of data. We break down why precision beats vagueness as social proof — and how to test it.
From the glossary
Vocabulary the article uses. See all terms →
- Shrinkage Estimator— A statistical adjustment that pulls noisy per-variant or per-segment estimates toward an overall average to reduce overfitting to randomness.
- Flicker Test— A diagnostic check to confirm whether users briefly see the original page before a variant renders in client-side experiments.
- Pre-Experiment Bucketing Audit— A review process confirming users are being randomly and evenly assigned to experiment variants before launch.
- Above-Fold Exclusion Rule— A test-design rule that deliberately prevents changes to above-the-fold elements so an experiment isolates below-the-fold effects.
- Flicker-Free Testing— A technical setup that prevents users from briefly seeing the original page before a test variant loads.
- Orthogonal Test Design— A test structure where multiple variables are changed independently so their individual effects can be measured without interference.
See more like this
ABWatcher catches A/B tests like this across 1,000+ brands.
Watch live experiments at 1,000+ high-converting brands, complete with hypothesis and takeaway. 7 days of Pro, free, no card, then Free forever for ten watched companies.