If we show only warm contacts (LinkedIn 1st-degree connections or prior interaction history) in the Growth Inbox, we will increase the Growth Inbox action rate from 5% to 15% among active BizDev users within the 14-day window.
Baseline (5%) from MongoDB: 54,371 Growth Inbox records, 5% action rate across all contacts. Target (15%) is 3× the current rate, supported by Sarah Mittiga’s observed behavior: she acts exclusively on contacts she is already connected to on LinkedIn. Segmented baseline (warm vs. cold action rates) should be queried from MongoDB before the experiment launches.
Three customer calls reviewed: Alex McGregor’s consulting team (Canada), Sarah Mittiga’s HR/consulting agency (Dubai), and Lisa Haydon of Pivotal Growth (Canada). All three confirmed the hypothesis. The variation is in how explicitly.
Alex McGregor
Consulting (Canada) — Early user
Immediately lit up at past-client resurfacing: “We get a lot of people we’d love to turn into repeat clients that for whatever reason we just don’t come back to.” Growth Inbox is not yet part of their workflow, but warm-contact re-engagement is the use case they want.
Sarah Mittiga
HR/Consulting (Dubai) — Active user
Already self-applying the filter. She checks “connected on LinkedIn” as her first filter before acting on any Growth Inbox card. She is not sending connection requests to strangers. She is acting on people she already knows. This is the behavior the feature would formalize.
Lisa Haydon
Pivotal Growth (Canada) — Power user
Called the newsletter-to-Growth-Inbox integration “freaking amazing” — a warm-contact feature by another name. Has 800 newsletter subscribers she is not systematically acting on. Her entire Growth Inbox usage is oriented around people who already know her work.
“At the moment I’ve always been having people that I’m connected to — so it says ‘connected on LinkedIn’ — so that means we’re already connected... here I would then maybe just say ‘add to prospects.’”
Sarah Mittiga — HR/Consulting, Dubai — strongest behavioral confirmation“This is freaking amazing. Because we don’t do that enough... you have to download [the list], it’s a pain in the ass, and it doesn’t pattern — it’s static for the month.”
Lisa Haydon — Pivotal Growth, Canada — on newsletter-to-Growth-Inbox integration“We get a lot of people that we’d love to turn into repeat clients that for whatever reason we just don’t come back to.”
Alex McGregor — Consulting, Canada — confirms the unmet needThe original hypothesis has been validated and sharpened in two ways by the discovery:
Warm contact + signal, not just warm contact. Sarah does not want a flat list of everyone she’s connected to — she wants to know when a signal occurs for someone she already knows (funding round, promotion, software change, disclosed timeline). The filter should surface warm contacts when there’s a reason to reach out. This sharpens the feature spec: it is not a static filter, it is a triggered warm-contact alert.
Volume reduction is a co-primary outcome. Sarah’s top stated frustration is that the Growth Inbox feels endless. She does not know when she is “done.” Filtering to warm contacts reduces cognitive load as much as it improves relevance. The experiment should track both action rate (relevance) and dismissal rate (effort reduction).
The MVP is a LinkedIn connection toggle. Gia already shows “connected on LinkedIn” on Growth Inbox cards (Sarah can see it). The fastest version of this feature is a UI filter: “show only contacts I’m already connected to.” No new data acquisition required. This can validate the hypothesis before the full signal-triggered version is built.
Discovery draws on two sources with different levels of interpretive strength. Both point in the same direction.
3 targeted discovery interviews. Calls designed specifically to elicit BizDev usage patterns. All three customers confirmed the warm-contact preference — one (Sarah Mittiga) provided direct behavioral evidence by describing a self-applied LinkedIn connection filter she already uses in the Growth Inbox. This is the primary evidence basis for the hypothesis.
516-transcript research database corroboration. Cross-referenced against the full Gia qualitative database. 105 of 272 distinct companies (39%) independently surfaced relationship-selling preferences, warm-contact language, or cold-outreach aversion in calls not designed to elicit this information. Several customers explicitly described the newsletter-integration feature by name. This is corroborating evidence — it shows the pattern is not isolated to the discovery interviews — but it is not a measurement of prevalence across the full customer base.
The 61% of transcripts in the database that do not surface this theme are structurally silent — sales and onboarding calls where BD strategy was not the topic. This absence is not disconfirming. It means the database can corroborate but cannot precisely measure how widespread the warm-contact preference is across all of Gia’s customers.
| Claim | Supported? |
|---|---|
| Consultants prefer warm outreach over cold prospecting | Yes — strongly, across both sources |
| Customers are already self-filtering Growth Inbox to warm contacts | Yes — direct behavioral evidence (Sarah Mittiga) |
| Newsletter subscribers are seen as the warmest leads | Yes — multiple customers, unprompted |
| Warm-contact preference is pervasive across all of Gia’s customers | Directional — not precisely measurable from this data |
| The filter will increase action rate from 5% to 15% | Not yet — the experiment will answer this |
The job of qualitative discovery is to justify building the experiment, not to predict the outcome. The evidence is sufficient to justify the build. The A/B test is what sizes the effect.
MongoDB baseline: 54,371 Growth Inbox records, 5% overall action rate, 10% open rate. Ignore reasons are never captured (field exists, always null). Data is not yet segmented by relationship warmth — that query is the next step before the experiment launches.
Before committing to build, Marty Cagan’s four-risk framework asks whether the product is valuable, usable, feasible, and viable. A product can fail on any one of these independently. Here’s where this experiment stands on each.
Mostly de-risked. Residual risk on behavioral effect.
The preference is validated: customers prefer warm outreach, multiple users have asked for this feature unprompted, and one customer (Sarah Mittiga) has already invented the behavior on her own. That de-risks the “do they want it” question.
The residual value risk is behavioral: will filtering to warm contacts actually cause users to act more, or will they simply see a smaller inbox and behave the same way? The discovery confirms preference; it does not confirm that preference translates into measurably higher action rate. That gap is exactly what the experiment tests.
Status: Sufficient to justify building MVP. Not sufficient to predict effect size.
Low risk for MVP. Higher risk for the full version.
The MVP — a toggle showing only LinkedIn 1st-degree connections — is a simple binary UI on an existing screen. Gia already displays a “connected on LinkedIn” badge on Growth Inbox cards, so the concept is already familiar to users. Toggle on, fewer cards, all warm. Low usability risk.
The risk rises with complexity. The full vision (warm contact + trigger signal, newsletter subscribers, past clients, closed-lost) introduces a multi-signal model that needs clear UI explanation. Base Pricing surfaced a real usability concern: “how do you know what’s a real connection and what’s not?” If LinkedIn 1st-degree is the warmth definition, some users will find that too broad; if it’s narrower (interaction history, engagement signals), users may not understand why contacts appear or disappear.
Status: De-risk the MVP in sprint 1. Prototype the multi-signal UI before committing to full build.
Low risk for MVP. Medium risk for full version. One open question.
The MVP depends on a single open question: does Gia have LinkedIn connection status on all Growth Inbox records, or only some? If coverage is full, the toggle is a filter UI with no new data work. If coverage is partial, the filter will silently omit warm contacts that aren’t tagged — which may produce a worse experience than the current unfiltered inbox for some users. Engineering needs to answer this before committing to the MVP build.
The full version requires new data ingestion pipelines — newsletter platform APIs (Wix, Mailchimp), interaction history matching, CRM sync. These are meaningful engineering investments and introduce third-party dependencies that are outside the team’s direct control. That build should not be scoped until the MVP toggle validates the behavioral effect.
Status: Answer the coverage question before committing. Do not scope the full version until MVP results are in.
One meaningful concern: perceived value reduction from volume drop.
If the filter reduces inbox volume significantly, some users may interpret a smaller Growth Inbox as Gia “doing less” — not as a quality improvement. This is a framing risk, not a product risk, but it is real. The toggle needs to be presented as a quality filter (“these are your warm contacts”) not a subtraction (“your inbox now has fewer items”).
The upside case is strong: if warm-filtered action rate increases materially, this becomes a retention and expansion driver — users who engage more stay longer and grow their seat count. Higher engagement per impression is a better product metric than high volume with low action. The business case for the MVP is sound.
The viability risk for the full version is scope and dependency. Newsletter integrations require partnerships or API agreements. CRM sync requires user setup and ongoing maintenance. These are not blockers but they are cost centers that need to be weighed against the revenue impact the experiment validates.
Status: Frame the filter as a quality upgrade, not a volume reduction. Validate revenue impact at MVP before investing in integrations.
Risk Summary
Value and usability risks are sufficiently de-risked to justify building the MVP toggle. Feasibility has one open question (LinkedIn connection coverage) that engineering must answer first. Viability is sound for the MVP; the full integration build carries scope and dependency risk that should not be committed to until the experiment validates the behavioral effect. The staged approach — toggle first, integrations after — is the right way to manage all four risks.
Two of three original blockers are supported. One remains.
Hypothesis validated. Three customer transcripts confirmed the warm-contact preference unanimously. Sarah Mittiga’s existing behavior is direct behavioral evidence. The hypothesis is ready to turn into an experiment design.
Baseline approach defined. Primary metric is Growth Inbox action rate (5% today). Target is 15%. Secondary metrics: dismissal rate and time-to-action. Segmented MongoDB query (warm vs. cold action rates by LinkedIn connection status) should be run to sharpen the baseline before launch.
Feature must be built. Gia does not currently offer a warm-contact filter on the Growth Inbox. The MVP — a toggle showing only LinkedIn 1st-degree connections — can be scoped and built before the experiment window. Full signal-triggered version (warm contact + trigger event) is the follow-on build after validation.