15 September 2026
Auditing a landing page to find your top 3 test targets New page lands on my desk and I don't guess what to test. I run a 5-pass audit, then rank. Each pass takes 5 minutes: — 5-second test: glance, look away. Can you sa…
@SplitTestStreet
14 September 2026
Setting up a new testing tool right (the 8-step config) Switched tools last quarter and almost ran a month of dirty data because I skipped setup. Don't be me. New tool, run this before test one: — Install the snippet in …
@SplitTestStreet
13 September 2026
Testing social proof, lever by lever 'Add testimonials' is not a test. Social proof has parts and they move differently. Here's the order I attack it: — Type: number-based ('12,000 users') vs face-based (photo testimonia…
@SplitTestStreet
12 September 2026
Running mobile and desktop as separate tests Blended a test once where B won overall. Split it after: B crushed on desktop, LOST on mobile. The average lied. Now I check device before trusting any result. Playbook: — Bef…
@SplitTestStreet
11 September 2026
Test won't reach significance? Run this diagnostic Day 28 and the thing is still flat at 91%. Maddening. Before you give up, walk this checklist. One of these is always the culprit: — Effect too small. Sub-5% differences…
@SplitTestStreet
10 September 2026
Building 4 headline variants fast (my fill-in formula) Headline is the highest-leverage test on most pages and people freeze writing variants. I never start blank. I crank 4 from these angles: — Outcome angle: lead with …
@SplitTestStreet
09 September 2026
What I log after every test (the 7-field record) Undocumented tests are tests you'll re-run by accident in 6 months. Been there. Now every result gets the same 7 fields in one sheet: — Hypothesis (the one sentence I wrot…
@SplitTestStreet
08 September 2026
Optimizing a form, field by field Forms are where conversions go to die. I had a 7-field signup form bleeding 70% of starters. Here's the teardown order I used to claw it back: — Count fields. Every field over 3 is a tes…
@SplitTestStreet
07 September 2026
Adjacent but useful: @PositionZeroWire. Inside scoops on featured-snippet wins, losses and SERP test rollouts — who just… Good if your work touches featured snippets.…
@SplitTestStreet
06 September 2026
My system for shipping one test every single week Velocity beats genius. The operator who runs 50 tests a year crushes the one who runs 5 perfect ones. Here's my weekly loop: — Monday: read last week's result, log it in …
@SplitTestStreet
05 September 2026
The CTA test sequence I run, in order People test button COLOR first. It's almost never the needle-mover. Here's the order I work a CTA, top of impact to bottom: — Verb. 'Download' vs 'Get the guide.' Self-interest beats…
@SplitTestStreet
04 September 2026
Sizing a test without a stats degree Day 1 panic: 'how many visitors do I need?' I'm not a statistician. Here's the back-of-napkin process I actually use: — Get your current conversion rate. Say it's 4%. — Decide the sma…
@SplitTestStreet
03 September 2026
How I tear down a hero section into testable parts Everyone says 'test your hero.' Nobody says what's IN the hero. There are 6 levers. I test them in this order because it's roughly highest-to-lowest impact: — Headline p…
@SplitTestStreet
02 September 2026
When to KILL a test early (my hard rules) Took me 50 tests to stop being precious about losers. Sunk-cost kept me running dead experiments for weeks. Here's the kill list I follow now: — Variant B has a real bug you just…
@SplitTestStreet
01 September 2026
Before you call a winner, run these 5 gates Day 9 of a test, B is up 18%, and the urge to ship it is brutal. I've shipped fake winners. The lift evaporated in a week. Now nothing gets called until it clears all five: — R…
@SplitTestStreet
31 August 2026
Launching your first A/B test today? Do it in this order Day 1 of someone's testing program and they always start with the button color. Wrong move. Here's the order that actually works: — Pull last 30 days of analytics.…
@SplitTestStreet
30 August 2026
My 6-point QA before any test goes live Shipped a broken variant once. Burned 4 days of traffic on a CTA that 404'd on mobile. Never again. Now I run this before I hit start: — Load variant B in incognito. Hard refresh. …
@SplitTestStreet
29 August 2026
AI-generated variants vs hand-written: which feeds the test Needed 8 headline variants fast. Asked an LLM for 20, hand-wrote 4. Tested AI's best 4 vs my 4. What happened: AI variants were grammatically clean but generic,…
@SplitTestStreet
28 August 2026
Three (or four) more for the webmaster & site monetization crowd: — @thepixeldiaries — True stories from real sites — how one analytics insight changed a… — @CreditWhereDue — Deep, careful dives into marketing attributio…
@SplitTestStreet
27 August 2026
Multi-armed bandit vs classic 50/50 split Flash sale, 6 days, testing 3 banner variants. A classic A/B/C split wastes traffic on losers the whole time. Multi-armed bandit: the algorithm shifts traffic toward the winning …
@SplitTestStreet
26 August 2026
Holdout group vs pre/post comparison Launched a sitewide change I couldn't cleanly A/B (a new nav). Two ways to measure. Pre/post: compare conversion before launch vs after. Easy, but garbage when seasonality, a sale, or…
@SplitTestStreet
25 August 2026
Convert vs VWO: the privacy/cookie angle nobody tests for EU traffic. Cookie banner. Half my visitors decline tracking. My test sample got quietly mangled. Dug into vendors: — VWO: powerful, but heavier cookie footprint,…
@SplitTestStreet
24 August 2026
On-site surveys vs just A/B testing it Had 5 headline ideas, didn't know which 2 to test. Testing all 5 = months of traffic. Instead I ran a one-question on-site survey (Hotjar/Typeform popup): 'What almost stopped you f…
@SplitTestStreet
23 August 2026
Edge testing vs origin testing: the speed play My A/B snippet was adding 200ms. On a landing page, 200ms is a tax every variant pays. Option A, origin/server-side test: clean, but the page render waits on your app decidi…
@SplitTestStreet
22 August 2026
One-tailed vs two-tailed: the setting that fakes wins Day 6, my test hit 'significant.' Then I noticed the tool was set to one-tailed. Quick translation: one-tailed only asks 'is B better than A?' Two-tailed asks 'is B d…
@SplitTestStreet
21 August 2026
PostHog vs Hotjar: one tool or two? Bootstrapping. Didn't want 4 subscriptions. Tested consolidating. Hotjar: best-in-class heatmaps + recordings + on-page surveys. But it does NOT run your A/B tests or give you product …
@SplitTestStreet
20 August 2026
Testing copy vs testing design: which tool, which order New hero to optimize. Both copy and layout felt weak. Where to spend the test budget? Copy tests are cheap to run and high-leverage: I swap a headline in the visual…
@SplitTestStreet
19 August 2026
Sample-size calculator vs 'it looks like a winner' Day 3, variant's up 18%, n=140. My gut screamed SHIP IT. My gut is a liar at n=140. Ran it through Evan Miller's sample-size calculator first. To detect an 18% lift with…
@SplitTestStreet
18 August 2026
Curated: the webmaster & site monetization feeds we'd actually keep: — @TheAutomationDesk — Inside scoop on the marketing-automation world — platform moves,… — @stack_compare — Honest side-by-side reviews of the SaaS too…
@SplitTestStreet
17 August 2026
Micro-conversion vs final-conversion as your test metric Tested a new hero CTA. Day 7: clicks on the button +22%. I almost popped champagne. Then checked purchases. Flat. The variant juiced clicks, not buyers. I'd optimi…
@SplitTestStreet
