A consumer AI writing platform with student adoption to spare — and a mandate to sell to the institutions those students already belong to. Here's the B2B engine underneath it.
Jenni.ai grew up B2C. Students found the product on their own — through search, through TikTok, through a professor's off-hand recommendation, through the friend in the next carrel who was already using it. Adoption compounded from the ground up, one paper at a time.
The pivot the business needed was the one every consumer SaaS company eventually stares down: the same seat, sold institutionally. Not one student at a time — but the whole university, the whole department, the whole research group, on a team plan negotiated with the person who signs the software contracts. That person is not the student. It's a Dean, a Provost, an Academic Director, a CIO — and they don't buy the way a 22-year-old buys.
When iExcel picked up the engagement, the B2C playbook was already firing. What didn't exist yet was the B2B acquisition engine underneath it: audience layer, persona segmentation, cold outbound, LinkedIn paid, CRM architecture, attribution. The product was ready. The go-to-market motion was not.
The engagement opened with a full strategic B2B marketing audit and acquisition proposal — a digital-marketing and go-to-market assessment scoped against where Jenni.ai actually was, not where the B2C growth deck assumed it was. Analytics footprint. Attribution. Site funnels for the institutional visitor. The overlap between the B2C engine and the B2B one — and, more importantly, everywhere they diverged.
Universities buy differently than students do. The visitor lands on a different page. They read a different proof point. They forward the link internally, then wait weeks for a committee to weigh in. None of that is what a B2C site is built to serve. The audit named the gaps, priced the fix against the impact, and turned the proposal into a phased build-out the CEO could stand behind — audience first, then LinkedIn, then outbound, then the CRM spine that would carry all of it.
Every phase downstream inherits the assumptions the audit set. When those assumptions are wrong, the campaigns get built on drift. When they're right, everything after compounds.
You cannot sell to a Dean and a CIO with the same copy. You cannot pitch a Provost in São Paulo the same way you pitch one in Riyadh. The B2B build-out started, as it always does, with the audience layer — sourced, verified, and cut against the vectors the business would actually campaign against.
Five language and geography groups, each with its own list build, its own tone, and its own inbox behavior. English core across US and APAC; Spanish, Arabic, and Portuguese for regional depth.
Academic decision-makers, split by title, discipline, and institution type. Public research universities. Private liberal-arts. Community-college systems. Graduate schools. Each one a different sales motion.
The technology decision-maker inside the same institution — a parallel track. Different pain, different pitch: procurement process, data policy, integration with existing academic-integrity tooling.
Not every persona signs the contract — but every persona shapes the deal. Research Associates, Academic Officers, and faculty leads flagged as influencers, campaigned differently from budget-holders.
Every academic institution is really five overlapping buying committees, wearing the same lanyard. Each persona reads a different subject line, opens a different attachment, and forwards the pitch to a different colleague. The persona matrix below is the map iExcel campaigned against — one message architecture per role, one CTA per role, one lifecycle path per role.
The paid layer had to serve the segmentation model, not fight it. LinkedIn Ads was the channel where the academic personas were actually reachable — where a Professor in Manila, a Dean in Madrid, and a CIO in Cape Town could each be matched against a title-and-institution filter that Meta and Google couldn't replicate.
The campaign architecture followed the persona matrix directly. One campaign set per role. One campaign set per geography inside that role. Copy re-cut for the pain point, the lever, and the CTA that persona actually converted on. The winning segment, by margin, was the Professor — and inside that, APAC Professors delivered the highest-efficiency acquisition of the entire campaign set. That efficiency wasn't incremental: the top APAC Professor segment landed leads 82% below the blended cost of the full campaign set.
The B2B campaign set didn't need to be complicated. It needed to be segmented. Once the segments were right, the budget stopped being spread thin and started stacking against the roles that were actually converting.
One 30-day window across the segmented B2B campaign set. Later 30-day windows produced 1,050 in a comparable read. The segmentation model — persona × geography × language — was the compounding lever.
Cold outbound is a discipline problem before it is a copy problem. The infrastructure went in first — Instantly and Lemlist for sending, Clay for enrichment, deliverability paced against warm-up rules, every list run through verification before it fired. Nothing in the outbound stack was allowed to run against an audience that hadn't been cut through the persona matrix.
Once the plumbing was clean, the copy layer went in. Persona-specific messaging by title, geography, and language. Sequences tuned per role — a Dean opens a different subject line than a CIO opens. Later phases layered AI-personalized outbound systems on top: enrichment signals pulled into the message, institution-specific angles auto-generated, so a lead about "AI writing for your composition department" landed differently in a research university than in a community college.
Persona-tuned copy showed up in the engagement numbers. The primary campaign into academic deans opened at 76% — roughly 2× Mailchimp's published education-industry benchmark — while a parallel Spanish-language dean sequence converted clicks at 23%, about 7× that same benchmark.
The result wasn't a spray-and-pray outbound program. It was a segmented, verified, persona-tuned engine that respected the reader's role before it asked for the meeting.
None of the above holds up without a CRM that can catch it. iExcel scoped the HubSpot architecture Jenni.ai needed to sit underneath the acquisition engine — lifecycle stages that reflected the academic buying committee, custom properties for persona × geography × institution type, source attribution wired to LinkedIn Ads, Instantly, Lemlist, and the AI-personalized outbound systems above.
The tracking and reporting layer sat on top of it. What was working in APAC. What was working with Deans. What was working when the enrichment signal was strong versus weak. The dashboards leadership needed weren't the ones a consumer product had ever built — this was B2B pipeline reporting on top of B2C acquisition velocity, and both had to coexist without one obscuring the other.
The result was a system of record the sales team could actually run against — and a leadership view that stopped being "what does the ad account say this week" and started being "which persona × geography combination is compounding, and where does the next dollar go."
Architecture planning against the academic buying committee. Lifecycle stages, custom properties, and dashboards scoped for B2B on top of the B2C velocity underneath.
Segmented campaign set — one per persona × geography. APAC Professor as highest-efficiency segment. Matched audiences fed back into HubSpot with source attribution intact.
Cold-email + LinkedIn automation stack. Deliverability-paced, verification-gated, persona-tuned sequences. Later layered with AI-personalized message generation.
List sourcing and enrichment. Every academic title, institution type, and geography vector filled in before the record entered a sequence.
Later-phase layer on top of Instantly/Lemlist. Enrichment signals fed into message generation so institution-specific angles landed automatically inside each sequence.
Persona × geography × institution type reporting on top of HubSpot. Leadership could see which combination was compounding — not just which channel spent.
If you're running a consumer SaaS with strong bottom-up adoption — and you're staring at the same seat, sold institutionally, as the next order of magnitude — the Jenni.ai engagement is the shape of what to do about it.
Audit the ground. Build the audience layer against the persona matrix, not against a single "customer." Turn on the paid channel where the persona actually is — for academia, that's LinkedIn, and inside LinkedIn it's often APAC. Layer outbound on top with persona-tuned copy and deliverability-clean infrastructure. Wire the CRM spine so leadership can read the persona × geography breakdown, not just the spend.
The B2C engine is not the enemy of the B2B engine. Built right, one becomes the demand-generation flywheel that feeds the other — students inside the walls of the institution the sales team is trying to sell to.
A B2B audience layer against your persona matrix. LinkedIn segmented by title × geography. Outbound tuned by role. A CRM spine leadership can actually read. Book a 15-minute sanity check on your pivot — we'll tell you what's ready and what isn't.
Book a 15-min sanity check →