LightbulbOne GTM platform. Every AE over quota.
Maple sells AI phone agents to 1,000+ restaurants. Lightbulb embedded as Maple’s RevOps and GTM systems engineering partner in May 2026 and built one platform where the whole revenue team works, with Salesforce, billing, the dialer, call data and Slack wired together and kept running in production. In Q3, every Maple AE finished between 120% and 150% of quota.
- Client
- Maple, AI phone agents for 1,000+ restaurants
- Partner since
- May 2026
- Our role
- Embedded RevOps and GTM systems engineering partner
- Systems
- Salesforce, Postgres, Stripe, Sequence, Nooks, Slack
176 pull requests to production
Every change shipped as reviewed code, deployed through CI next to Maple’s product, across every part of the revenue lifecycle.
Automations run on their own: syncs, jobs, Slack alerts, AI pipelines.
Features and tools are what reps and leaders use: pages, fields, reports and workflows.
Fixes and maintenance keep it all running as the business changes.
Reps don’t log into Salesforce
Maple’s sellers work in one place: the GTM app. Behind it, the app reads and writes Salesforce, Postgres, Stripe, Sequence, the dialer and Slack, so Salesforce stays the source of truth without being somewhere reps have to go. We built into that platform and maintained every integration around it, shipping 176 pull requests to production since May.
GRR and NRR dashboard
Gross and net revenue retention and cohort retention, compiled from Stripe and Sequence billing data and split enterprise vs self-serve.
AI email sequencer for SDRs
A call outcome in the dialer triggers a personalized follow-up email. The SDR reviews each draft in the loop before it goes out.
AI on every activity
Calls, meetings and emails run through LLMs and decision models like Jev, which fill in CRM fields automatically.
Quota and commission dashboard
Synced to Stripe and Salesforce: billed revenue, eligibility windows and accelerators. Reps work deals instead of tracking commission by hand.
Automations page
A Zapier-style view inside the GTM app of every backend automation, with the Slack notifications it sends and the Salesforce writes it makes.
Two-way sync, maintained
Salesforce and Postgres kept in step, plus live integrations with billing, the dialer, phones, calendar and call transcripts.
In Q3, every Maple AE hit quota
A platform only matters if the numbers move. We used it to run Maple’s revenue engine: one dialer, a deal score leadership can forecast from, a deal cycle that captures the right data, and a steady supply of fresh accounts for every rep.
An AI score that tells leadership which deals are real
Call transcripts update a 1 to 5 deal temperature on each opportunity. Deals scored Hot before they closed won 84% of the time, and not one deal scored Likely Lost was won. When the score marked a deal as cooling (Neutral or lower), it was lost 97% of the time. Leadership forecasts from it, and reps use it to decide where to spend the week.
Every demo fills in the deal record
We reshaped Maple’s stages around exit criteria, then had AI fill the record from every call: next step, objections, blockers, pain, timing, competition and deal temperature. A deal that reaches a demo now averages 4 of these 8 fields filled, up from half a field before. That’s 8 times the deal data, with no typing from the rep.
From Salesfinity to Nooks
We moved the SDR team to Nooks, consolidating dialer spend and cutting the dead air between parallel dials. Every dial since the switch writes its outcome back to Salesforce.
Fresh lists, clean books
We kept new lead lists flowing and cut books for every AE and SDR, with duplicate restaurants cleaned out across territories.
SDRs call the right restaurants. AEs skip the admin.
The last step was time. We gave SDRs a model that ranks who to call next, and took post-call admin off the AEs’ plates.
A predictive model that ranks every account each morning
Every morning the model scores each account in the SDR books and tells each SDR who to call next. It was trained on Maple’s own call outcomes and validated on held-out weeks before launch. Once it went live, SDRs booked 13 demos in a day against an average of 8.5.
- 01Call or demo endsTranscript arrives from the dialer, phone or meeting recorder
- 02AI drafts the updatePain, timing, next step, objections, competitor, deal temperature
- 03Rep reviews in SlackBlock Kit cards for the draft and the meeting outcome
- 04Salesforce updatedWritten back automatically, with history
Want a GTM platform your reps actually use?
Lightbulb is the technical systems and data layer for revenue teams. Start with a free CRM and GTM Systems Diagnostic.