A flashcard platform hit its internal SEO ceiling. iExcel rebuilt the discoverability layer — URLs, robots, sitemap, titles, keyword map.
Brainscape is a flashcard-based learning platform grounded in cognitive science — spaced repetition, active recall, and confidence-based intervals — sold across the web and an iPhone app. The tagline that lived on the homepage for years said it plain: what would you like to learn today? Millions of study sessions, thousands of subjects, and a deep bench of user-generated decks sat behind that door.
The catch: Brainscape had built discovery on its own product taxonomy — nested market URLs, category-then-subject paths, and a homepage title that told a story to humans without saying anything to Google. Growth from search had flattened. Every subject page competed for the same generic phrase. Blog posts weren't in the sitemap the platform actually served. Robots.txt was silent on directories that were burning crawl budget.
The team knew there was a ceiling. They asked iExcel to identify it, name it, and hand back a plan to break it. Not a rebuild. A restructure.
The single biggest lever on the audit was URL structure. Brainscape's market pages nested subjects under category, sub-category, and subject — a clean data model, a poor discovery surface. Search engines were being asked to weight market and category1 as if they carried semantic value. They don't. The keywords that mattered — spanish flashcards, french flashcards, ap flashcards — were buried three segments deep in a URL that read like a folder tree.
The recommendation was a keyword-forward URL convention. Move the phrase people actually search into the URL. Kill the nested prefix that meant nothing to a crawler. Give each premium subject a landing page that reads its intent at the URL level — before a title tag, before an H1, before a single word of body copy is parsed.
Robots.txt was the second lever. Brainscape's file was thin — the platform was letting crawlers waste budget on system paths that had no business ranking. Assets. Cart. Sign-in. Sign-up. Teacher tools. Profile pages. Market-screen previews. None of it should have been indexed. All of it was being crawled.
iExcel wrote a custom robots.txt covering the crawlers that actually matter to a US-first EdTech: Googlebot, Bingbot, msnbot, Slurp, and Googlebot-Mobile. Every non-indexable directory disallowed. Crawl budget redirected toward the pages that could rank.
The sitemap had a second problem. Blog posts — the exact content Brainscape spends effort producing — were not in sitemap.xml. Either fold the blog into the main sitemap or serve a dedicated blog-sitemap.xml. Either way, tell Google the content exists.
Googlebot, Slurp, Bingbot, msnbot, and Googlebot-Mobile — the crawler set that carries US-first EdTech search traffic across desktop and mobile./assets/, /cart/, /images/, /market_screens/, /market_screens_thumbs/, /packs/, /profiles/, /sign_in/, /sign_up/, /teachers/. System paths, thumbs, gated flows — none of it belongs in an index.sitemap.xml. Recommendation: fold blog URLs into the primary sitemap, or serve a separate blog-sitemap.xml referenced from robots. Either way, the content stops being invisible.Fixing discoverability only matters if the page holds attention once a visitor lands. Core subject pages — the standardized-test and foreign-language landing pages carrying the platform's highest-intent search traffic — were bouncing at a rate well above what a B2C site should tolerate. The landing experience lacked the keyword alignment and on-page structure to convert the demand the technical fixes were about to unlock.
"What would you like to learn today?" is a great tagline. It's also a title tag that tells Google nothing about what Brainscape does, who it's for, or what subjects the platform actually covers. Zero keyword surface. Zero subject signal. A homepage title doing brand work — and no discovery work.
iExcel rewrote both the title and meta description against the exact subject queries the keyword map surfaced. Brand still leads. Category is named. The premium subjects — the ones that carry search volume — are now on the page in the elements Google reads first.
The keyword workstream came out of a full Google Keyword Planner pass, then organized into ad-group-style clusters the marketing team could actually build against. Every row tied to a subject. Every subject tied to a landing-page recommendation. Every premium topic got its own row — and its own URL slug in the restructure.
The largest volume cluster. Language flashcards read as brand-agnostic long-tail — high commercial intent, clear landing-page anchor per language.
The exam-prep cluster. Test-prep intent, deep purchase signal, and Brainscape's spaced-repetition product speaks directly to it.
The academic-subject cluster. Undergraduate and high-school demand, each subject earning a dedicated landing page in the restructure.
The specialty-vertical wedge. Small volume, low competition, defensible click-share — the kind of niche that pays off SEO discipline.
The differentiator cluster. Terms that tell the memory-science story Brainscape actually delivers — top-of-funnel education queries.
The product-category cluster. High commercial intent, direct competitor overlap — worth the on-page work and the ad-group focus.
The audit didn't stop at fixes. Every recommendation came with a principle the Brainscape team could apply to pages iExcel hadn't touched. Words versus exact-match. Headers versus body copy. Hyphens versus underscores. Mobile versus desktop parity. The rules that make SEO work in 2020, wrapped so the internal team could keep executing after the engagement wrapped.
Individual keyword tokens still carry weight — even when the exact phrase isn't hit. Optimize for keyword coverage first, exact-match second.
H1 and H2 tags weight harder than body prose. Every premium subject page gets its target keyword in the heading — not buried in a paragraph.
Kebab-case — spanish-flash-cards, not spanish_flash_cards. Hyphens split into tokens; underscores don't. It's a one-character decision that shapes readability.
Repeat the target phrase enough that Google reads intent — not so much that the page reads like it's writing to a robot. Balance is what wins.
Mobile-first indexing means the mobile render is the render Google reads. Every recommendation gets applied to both surfaces — no desktop-only optimizations.
Every high-value keyword deserves a dedicated landing page — with a keyword-forward URL, a titled H1, a purposeful meta description, and content the crawler can actually rank.
The foundation for the 192-row map. Monthly volumes on 100 target terms — organized by ad-group topic so paid and organic can share a keyword spine.
The four elements Google reads before it reads anything else. Every high-value page audited against all four. The homepage rewrite in Chapter 04 is the template.
The keyword map organizes by topic — Spanish, French, AP, SAT, chemistry, brain science. One cluster, one landing page, one intent. Not a pile of orphaned keywords.
If you run a B2C platform — an app, an EdTech, a marketplace, a directory — with a strong product, a growing user base, and organic traffic that has quietly plateaued: the Brainscape audit is the shape of the diagnosis.
It usually isn't the content. It's the structure underneath the content. URLs built for the database, not for search. Robots files that let crawlers wander. Sitemaps missing the pages your team spent months producing. Homepage metadata written for humans that says nothing to Google. A keyword map that lives in someone's head instead of a spreadsheet the marketing team can build against.
You don't need to rebuild the product to break the ceiling. You need to rebuild the discoverability layer around it. Foundation first. Then the growth.
A URL structure a crawler can read. A robots file that directs the budget. A sitemap that includes the blog you actually publish. A homepage title that names the subjects you actually teach. Book a 15-minute sanity check — we'll tell you what's under-built and what to do about it.
Book a 15-min sanity check →