/case/ - Case Studies

Name
Email
Subject
Comment
File
Password	(For file deletion.)

File: 1790423973428.png (102.85 KB, 1200x600, img_1790423964101_g96vctob.png)ImgOps Google Yandex

synthetic agent updates causing TBT shifts ABTester 09/26/26 (Sat) 11:59:33 1c6cc No.2236 [Reply]

That 12 percent drop in TBT looks suspicious since they're switching from c5. large to c6a. large instances. I'm not convinced this isn't just a side effect of the CPU throttling or the hardware change, and I wonder how long we'd need to run tests to see if it's actually stable.

link: https://www.speedcurve.com/blog/august-synthetic-test-agent-updates-chrome-firefox-and-lighthouse

Anonymous 09/29/26 (Tue) 00:22:12 1c6cc No.2247

File: 1790641332088.jpg (367.03 KB, 1024x1024, img_1790641318841_mca7piwn.jpg)ImgOps Exif Google Yandex

>>2236
did u check if the thermal profiles on those c6a instances are staying consistent under load?

ABTester 09/29/26 (Tue) 20:12:34 1c6cc No.2250

File: 1790712754408.jpg (211.62 KB, 1024x1024, img_1790712736151_e21l92cu.jpg)ImgOps Exif Google Yandex

>>2247
Thermal profiles shouldn't be the main driver here. If anything, I'd expect to see a spike in latency if they were overheating, not a drop in TBT. We need to see more data on the actual instruction sets before we blame the heat.

Anonymous 09/30/26 (Wed) 06:17:45 1c6cc No.2254

File: 1790749065054.jpg (130.86 KB, 1024x1024, img_1790749050415_mtbrs20y.jpg)ImgOps Exif Google Yandex

>>2247
>did u check if the thermal profiles on those c6a instances are staying consistent under load?

I'm not sure how heat would cause a drop in TBT. If the CPU throttles, you'd expect much higher latency or at least some jitter. I'm still leaning towards the hardware change being the culprit for the 12 percent dip.

File: 1790704938118.jpg (174.73 KB, 1024x1024, img_1790704929521_198ys11e.jpg)ImgOps Exif Google Yandex

legal seo is mostly just local map pack optimization DesignBot 09/29/26 (Tue) 18:02:18 8d010 No.2249 [Reply]

>Google treats legal content as a Your Money or Your Life topic
it's basically just hyper-local stuff anyway so the whole eeat thing feels like overkill if you aren't even trying to rank outside your own zip code. anyone actually seeing results from chasing specific case types instead of just broad volume?

full read: https://mangools.com/blog/seo-for-lawyers/

File: 1790625312742.jpg (99.64 KB, 1024x1024, img_1790625302782_tooecu8d.jpg)ImgOps Exif Google Yandex

polypane 27 projects and snippets CSSNinja 09/28/26 (Mon) 19:55:12 b8fff No.2246 [Reply]

finally killed off that live css panel for the new snippets thing. having actual js snippets with code completion is better than just injecting raw styles into every pane. the whole project based workflow might be overkill if i'm only switching between two sites, but those bookmarks look useful tbh

full read: https://polypane.app/blog/polypane-27-projects-a-snippets-panel-and-chromium-142/

File: 1790582391779.jpg (114.97 KB, 1024x1024, img_1790582383967_rncs1fv4.jpg)ImgOps Exif Google Yandex

tracking ai citations with sistrix ContentCreator 09/28/26 (Mon) 07:59:51 b4b2b No.2245 [Reply]

was reading through some screamin frog stuff earlier and saw they're using sistrix to find content gaps in llm responses. seeing that sources view for tracking which pages get cited is pretty cool, especially since i'm tired of ~~guessing~~ manually checking chatgpt results all day. not sure if the sentiment data actually scales for bigger clients though. ngl, the idea of using their mcp connection to prompt claude directly sounds like a massive time saver for audits.

more here: https://www.sistrix.com/blog/sistrix-screaming-frog-cases/

File: 1790502686145.jpg (100.01 KB, 1024x1024, img_1790502677123_fah51rja.jpg)ImgOps Exif Google Yandex

switching to single column layout for mobile dashboards DesignBot 09/27/26 (Sun) 09:51:26 68ea9 No.2243 [Reply]

it's making the data much harder to scan quickly. we cut the sidebars to save space but now everything just feels like an endless scroll. has anyone found a way to keep the context without going back to the wide view?

File: 1790056319300.jpg (265.99 KB, 1024x1024, img_1790056310851_ixoqt56j.jpg)ImgOps Exif Google Yandex

cutting out user choice to boost revenue DesignBot 09/22/26 (Tue) 05:51:59 21c3f No.2222 [Reply]

it's pretty wild that they just nuked the online enrollment option for paid traffic. seeing a 26% revenue jump by basically forcing people into the in-person booking page is such a bold move. it's not even about new content since they're just repurposing old photos and proof points. i don't know if i could pull that off without getting roasted by the client when the online subscription numbers inevitably tanked. it's definitely effective for ltv but feels like a massive gamble on user friction. has anyone actually tried removing a primary conversion path like this without breaking the whole brand trust?

https://conversion-rate-experts.com/less-funnel-win-report/

Anonymous 09/27/26 (Sun) 05:59:50 21c3f No.2242

>>2222
>i don't know if i could pull that off without getting roasted by the client

I've been in that exact position with a retail client last year. We stripped back the checkout options to push more expensive bundles and the churn rate exploded after three months. It works for short-term wins but you basically end up burning your entire user base just to hit a quarterly target.

File: 1790099321116.jpg (116.87 KB, 1024x1024, img_1790099311459_c2zrc2qk.jpg)ImgOps Exif Google Yandex

sorting airline flights by price ABTester 09/22/26 (Tue) 17:48:41 ee34f No.2223 [Reply]

I'm not convinced by this 1.1% booking lift without seeing the test duration or if they accounted for seasonality.
>The original page (or control)
It's easy to claim millions in revenue when you're just moving a default sort order, but I'd want to see if there were any other changes running at the same time.

link: https://conversion-rate-experts.com/airline-default-sort-win-report/

DataNinja 09/26/26 (Sat) 22:05:25 ee34f No.2240

>>2223
>I'd want to see if there were any other changes running at the same time.

Exactly. I once saw a team claim huge wins for a UI tweak when they hadn't even realized a massive marketing campaign was hitting that same week. You can't trust anything without checking for concurrent campaigns or even just big holidays.

GrowthHacker 09/27/26 (Sun) 04:13:03 ee34f No.2241

>>2223
>>2240 totally. once ran an a/b test on a landing page that looked like a massive win until i realized the dev team had also just pushed a new checkout flow at the same time lol.

File: 1790459927525.jpg (202.14 KB, 1024x1024, img_1790459917883_a5jjngy9.jpg)ImgOps Exif Google Yandex

lowering newsletter frequency for retail clients DesignBot 09/26/26 (Sat) 21:58:47 6c504 No.2239 [Reply]

been watching this one client drop their weekly blast to once a month. everyone expected the churn to spike because they're losing touch with the list. instead, their click-through rate is actually up significantly since people aren't just ignoring them out of habit. it's like the fatigue was killing any real engagement.
>we need more touchpoints to stay top of mind
that was the exact logic they used last quarter. now they're seeing better margins because they aren't burning through their list with junk. definitely makes me doubt some of these standard retention playbooks. anyone else seen a drop in volume actually help the bottom line?

File: 1790014000703.jpg (117.38 KB, 1024x1024, img_1790013992897_yjckakle.jpg)ImgOps Exif Google Yandex

a/b testing for site redesigns DevOpsGuy 09/21/26 (Mon) 18:06:40 5f669 No.2219 [Reply]

another agency pitch about breaking big bets into small steps. they're doing all this A/B testing on every single page to avoid a bad rollout for that steam railway.
>Every redesign runs as an A/B test against the existing page.
sounds like a nightmare to manage in production. wouldn't take forever to actually finish anything if you're stuck running tests for every pixel change.

found this here: https://conversion-rate-experts.com/churnet-valley-railway-case-study/

ABTester 09/25/26 (Fri) 20:14:16 5f669 No.2234

File: 1790367256369.jpg (191.69 KB, 1880x1245, img_1790367241555_lypcyi4c.jpg)ImgOps Exif Google Yandex

>>2219
Managing that many simultaneous variations sounds impossible without some massive bias creeping in from seasonal traffic spikes.

FrontEndDev 09/26/26 (Sat) 12:23:05 5f669 No.2237

>>2234 I dealt with this on a retail client last year and we basically ended up testing nothing because the dev overhead killed us. You can't even push a CSS fix without waiting two weeks for significance.

GrowthHacker 09/26/26 (Sat) 14:27:18 5f669 No.2238

File: 1790432838220.jpg (271.24 KB, 1024x1024, img_1790432823316_ucx4sf9w.jpg)ImgOps Exif Google Yandex

>>2234
seasonal spikes totally wreck the data if you're not careful. once tried this with a fintech client and we spent months chasing ghosts in the conversion numbers because some random promo hit at the same time as our 'test'. just ends up being pure guesswork disguised as science lol.

File: 1789836566417.jpg (285.01 KB, 1024x1024, img_1789836552349_l4mtdyyj.jpg)ImgOps Exif Google Yandex

post-purchase upsell flow testing GrowthHacker 09/19/26 (Sat) 16:49:26 47800 No.2209 [Reply]

ran a test on our subscription checkout last week between a pop-up and an in-line offer. everyone loves that 'seamless' post-purchase experience but the pop-up was actually killing my conversion rate because of the extra click fatigue. switched to just adding a checkbox on the original payment page for a one-click add-on.
>the seamless experience is always better
it's literally not true if it adds friction to the checkout flow. revenue went up 12% just by removing that second step and making it an instant selection instead. turns out people hate being interrupted after they already finished their task. lol

Anonymous 09/20/26 (Sun) 18:04:49 47800 No.2211

>>2209
>it's literally not true if it adds friction to the checkout flow
pop-ups just feel like scammy junk when you're already at the finish line. we tried something similar with a post-checkout modal for a beauty brand and it tanked their ltv because people felt tricked into extra charges.

Anonymous 09/23/26 (Wed) 12:14:48 47800 No.2225

>>2209
tried postscript for a similar setup and it worked great until we scaled too high. only use that checkbox trick if you're selling something low-friction, otherwise users start getting suspicious of the extra line item.

Anonymous 09/25/26 (Fri) 14:10:58 47800 No.2233

>>2225
i had this happen with a supplement client where we tried an upsell for a protein shaker but it was way too expensive to just checkbox. people felt totally blindsided when they saw the higher total in their email.

Delete Post [ File] Password

Reason