How RevQ became an account operating system.

Aug 30, 2026

When nothing happens, the app now tells you why

A blank screen and a silent failure look identical. Every surface that used to show one now names the actual cause.

• A provider that refuses a request is quoted in its own words. A payment hold on a data provider was being reported as an invalid API key, which sends you to rotate a credential that was never wrong.
• A phone number gets bought when your sequence needs one, even while a provider is unavailable. The rule is your campaign's channels, not which vendor happens to be up: if WhatsApp is a live step, the number is worth buying from whoever can supply it.
• Drafting works on a freshly accepted LinkedIn connection. That conversation has no messages yet, which used to return nothing at all; it now opens with a first touch written from that person's seat.
• A draft can no longer congratulate you on something the company did. Who a fact belongs to is treated as part of the fact, so a company's leadership change never becomes a stranger's new job.
• Lead searches apply the description you typed. A size and location brief was being compiled into an empty filter and silently ignored, so the search ran wide; one search went from seven companies to eighty-six once it was honoured.
• The runs page opens on the last day something happened rather than an empty today, shows the day at a glance, and states what held the queue back. On one campaign that was a single rate-limited LinkedIn account holding ninety-six touches.

Aug 23, 2026

The committee became the execution boundary

RevQ now treats committee coverage as the operating definition of account readiness, from score to sequence to reply.

• A qualified account enrolls only people who fill a committee seat. If no one is seated, nobody enters cadence; if the sequence requires several channels, each person must be reachable on all of them.
• Tier now combines fit, live timing, and actionability. A high score with nobody reachable cannot present itself as the account your team should work first.
• Every touch is drafted from the person's seat, company facts, why-now evidence, relevant capabilities and proof, and the conversation already established with that person.
• A reply can add a referral, open a missing seat, pause or move cadence, update the account opportunity, and book or reschedule a meeting while preserving the full thread memory.
• Delivered projects now feed source-tagged, audience-specific case-study decks with plain-language revision, visual verification, approval, and PDF export.

Aug 1, 2026

Everyone on the committee, and what each account cost

Research went all the way for every account, but not for every person on it. Now a seat on the buying committee is reason enough to finish someone.

• If a person holds a committee seat or is one of your contacts, every field your sequence needs gets filled. The old rule asked whether a number would make them immediately sendable and bought nothing if not. But a seat the sequence cannot run on is a seat that does nothing.
• A request a provider accepts but can never answer no longer blocks the paid fallback. One required revenue owner was found, requested, and left with no number because the request counted as “in flight” when nothing could ever arrive.
• Researching a long list stopped hitting a wall. Work now spreads across every LinkedIn account you have connected and reuses the employee list already gathered for a company, so a batch that used to stall around sixteen accounts completes.
• You can see what each account cost. Spend was always recorded per account and never shown; it now sits on the costs page by company name.
• The money is counted honestly: provider credits are priced from your real plan instead of counted as free, a batch total no longer sweeps in unrelated work that ran at the same time, and work belonging to no single account is shown as its own line rather than spread across yours to tidy the arithmetic.

Jul 31, 2026

Research finishes the job, and the record obeys you

Two half-measures ended: research that stopped short of a contact you could actually use, and corrections that providers quietly overwrote.

• Every account you research now goes all the way. A person it finds is unmasked to a real name and a verified work address instead of sitting at “found, not reachable”, including the accounts that could never earn the standing that used to unlock it.
• A person is bought once, not once per seat they hold. At a small company the founder is often also the approver, and each seat used to be invoiced separately.
• The finished-research notification names names: which accounts were researched, which people you actually paid to enrich, and what you were truly charged.
• Research writes itself into Activity: the score with its arithmetic, each person found, each person paid for, and one honest line per account for whatever the run could not do. It also reaches each person's own timeline, which is where “why is this person in my book?” gets answered.
• Your corrections now outrank every provider: in the record, in the lists, and at the moment a message is actually sent. Contact details are editable right where you read them, and every change is kept with its before and after.
• Importing a list stopped being a coffee break. A 200-row sheet lands in about a second instead of minutes, with the same duplicate and ownership reporting as before.

Jul 30, 2026

Every message is written to the seat that has to say yes

You define the buying committee once, on your business profile. That definition drives every word that goes out.

• One card per seat: the decision it owns, and the designations that typically hold it. That is what a buying committee means everywhere in the product.
• Whenever a pipeline is generated, every email, LinkedIn message, and WhatsApp touch arrives with a version per seat already written. The commercial leader's case study talks rep productivity; the medical affairs owner's talks MSL insights and KOL work.
• Sequences you built earlier catch up with one click: structure untouched, your own edits never overwritten.
• At send time each recipient simply gets their seat's version; a seat without one falls back to the base, so nothing you didn't configure changes.
• The prospect never sees the mechanics. No message mentions committees or roles, but drafts you review carry a quiet “written to” tag, so you always know which seat shaped them.
• Enrolling became one idea as well: you enroll an account, and everyone on it who can run your full sequence goes in with it.

Jul 30, 2026

The interface stops making you read walls

An evening-long design ruling, applied everywhere at once: no page is a stack of full-width fields.

• Settings, account, data providers, sending domain, committee views and the booking page all reorganized into columns. Short fields share a row, cards sit two and three abreast, and skinny panels became side rails.
• Form fields went borderless at rest: the warm fill alone says “type here”, and the edge appears when you reach for it. Every tab strip is underline-style, and section headings carry small, meaningful icons.
• Every save now tells you exactly what changed, or exactly why nothing saved. That immediately caught a mis-pasted mailing address that had been silently riding along.
• Empty pages teach instead of apologizing, and zeros stopped dressing themselves up as charts.
• Sending-domain setup tells the truth step by step: verified, propagating, or setup incomplete. “What happened?” unfolds the whole pipeline as a checklist that names exactly where things stopped. Bringing a domain you already own is a complete, copy-paste flow.

Jul 28–29, 2026

Your identity, your library, your channels

The workspace sorted out what belongs to a person, what belongs to the organization, and what belongs to a single campaign.

• Your inbox, calendar, sending identity, timezone and currency belong to you, not to a campaign. Connect once, and every campaign you work uses it.
• Onboarding became four steps that file everything where it belongs, invitations know who they're for, and verifying an email, resetting a password or accepting an invite all use a one-time code.
• Say which data providers and outreach channels you actually use, and the product keeps that promise: a channel you didn't declare never sends, and a provider you turned off never spends.
• Research can buy phone numbers cheapest-first, so a WhatsApp sequence can actually run. A contact you chose to put in a campaign always gets its number bought, because that membership was your decision.
• The account page tells the whole story and names its gaps: what's known, who has to say yes, why now, how the score was arrived at, and the full activity ledger.
• Pipeline & sequence became its own page, and reads like a plan.

Jul 24–27, 2026

The account gets researched like an analyst would

The unit of work changed from a contact to an account, and scoring stopped being a number handed down.

• Select accounts, press Research & score, and see the price before anything spends. RevQ resolves the company's website, reads it and the public news, and works the batch one account at a time with a live meter.
• Every buying signal you declared gets its own verdict with evidence and a source link. The score shows its arithmetic, so a 94 reads as profile + timing + adjustments, never as an assertion. Signals you declared that are absent stay on the page with the reason.
• The buying committee is mapped as seats with real alternates. On accounts worth pursuing, RevQ unmasks the people, including real names, verified work emails, and LinkedIn, then files them as contacts by itself.
• Coverage means people you can actually reach on a connected channel, not checkboxes. Confirming a seat is one click, kept separate from the machine's own confidence.
• Re-importing a list never duplicates an account you've already researched, and re-running research never re-buys a person it already knows.

Jul 22–25, 2026

Your documents become a capability library

Upload your decks, ICP docs, and messaging. RevQ turns them into structured evidence it can sell with.

• Products with per-product buying committees, case studies, proof points, and client evidence are extracted from your own documents and website, editable, and confirmed by you before anything uses them.
• Your buying signals become priced scoring rules: each one weighted high, medium, or low, and the weight is exactly what the scorer pays when the event is real at an account.
• Decks and outreach cite this library. An image on a client-facing slide is a claim, not decoration: evidence you own, or visuals built from your own sourced numbers, never stock theatre.

Jul 13–14, 2026

Setup, split into small pieces

Setup was one long wizard, and it created friction. It's now small standalone units, and lead intake finishes itself.

• Business profile, per campaign: paste your website and get a drafted description, proof points, and pre-filled targeting. Every campaign can tell a different story.
• Find leads and Add leads are separate, continuous surfaces. Imports run inline on the page, and afterwards the next step is stated plainly.
• Duplicates are visible: an import names the other campaigns already running the same people, so you never double-touch someone unknowingly.
• Companies imported without people fill themselves: Apollo pulls the buying committee automatically, with verified phone numbers arriving asynchronously.
• Faster, cheaper scoring at a fairer bar: a ~3× cheaper model, and the qualification bar moved from 80 to 50.

Jul 11–12, 2026

Judgment-first workspace

The interface grew up around one idea: the machine does the work, you make the calls.

• My Day: the dashboard opens with what needs you now (replies waiting, meetings today, leads to review), each one a click away. A quiet campaign reads as an empty list.
• Profiles that decide, not just display: every fact resolved once across Apollo, Tracxn, and Probe42, organized by topic in tabs.
• Next-best-action on every lead, computed the same way on every screen.
• Quick-view drawers to peek at a person or company without leaving the table.
• A live view of where every lead sits on the plan right now, with overdue touches in red.

Jul 9–10, 2026

The data platform

Lead data moved from ad-hoc blobs to a real, shared foundation.

• A platform-global directory: companies and people are canonical entities with evidence from every source, so a lead known to one campaign enriches the next.
• Pay only for what the rules need: enrichment is need-driven, cheapest-first, reuse-first, and only ever at the moment a lead enters.
• You decide when to spend: scoring and enrichment are manual on purpose. The machine never spends money on its own.

Jul 7–8, 2026

Costs, oversight, and a new name

The platform became something you can run as a business, not just a demo.

• Every rupee attributed: enrichment, scoring, drafting, discovery, web search, and email metered per member and per campaign, with cost-to-date and cost-per-meeting.
• An admin console for teams running several campaigns at once.
• Real error and metric monitoring, so nothing fails silently.
• The platform took its name: RevQ, the revenue queue that fills itself.

Jul 5–6, 2026

The Conversation Agent

Replies stopped being handled one message at a time.

• Reads the whole thread: classification and drafting consider the full conversation, not just the latest email.
• Remembers: persistent per-conversation memory, so it behaves like an SDR who knows the history.
• Decides once, acts on everything: one reasoning pass per turn drives every intent, validated across dozens of full lifecycles.

Jul 3–4, 2026

It books the meeting

The pipeline learned to close the loop end to end.

• Every engaged reply drives toward a booked call, proposing real open calendar slots and counter-proposing when you're busy.
• No-shows rebook themselves; reschedules never double-book; a cancel only touches the meeting in context.
• A complete reply taxonomy maps every inbound to the right pipeline move, with no gaps.

Jul 1–2, 2026

Real prospecting

The machine started finding your buyers, not just working a list you gave it.

• A discovery agent searches the web for companies that fit your ICP, each with a concrete reason to reach out now.
• It identifies the buying committee, dedupes, and rotates search angles, then prepares each candidate for your explicit Score decision.

Jun 30, 2026

It runs itself

The moment a campaign stopped needing a human to push it forward.

• The engine tick: a scheduled heartbeat that, for every active campaign, syncs replies, advances the cadence, sends what's due, and handles meeting logistics.
• Multi-channel cadence across email, LinkedIn, and WhatsApp, paced within each channel's safe limits, drafted per-prospect at send time.
• A pipeline and sequence builder that generates a tailored funnel and per-stage cadence from your business.

Jun 28, 2026

Foundation

Day one shipped the whole skeleton.

• Multi-tenant accounts with strict, database-level data isolation.
• A resumable setup wizard: connect your inbox and calendar, describe what you sell, define who you sell to and how you score them, design the pipeline and sequence, set your sending identity, and choose where leads come from.
• The first working cut of every core surface: leads, pipeline, inbox, meetings, dashboard, and activity, with real Tracxn/CSV ingestion and Google Calendar booking.