002 · Paid Media

What this is

The performance side of Bet 3: the paid search and LinkedIn accounts rebuilt from the ground up on the consolidated ad-group structure, bidding re-aimed from form fills to CRM-stage value, and a new-format testing rail — executed by Verto's paid growth specialist (SOW §3.2) inside the plan, the measurement, and the tooling Redis owns.

Goal: lower cost per lead while expanding into new paid formats — and move the paid scoreboard from clicks to pipeline (pipe-to-spend 1.1x → ≥1.8x).

Why now

Most of the marketing budget flows through paid, which makes it the fastest place discipline pays — and the H2 memo says the quiet part: program-level return has historically run below break-even, with one audit line showing 91% of paid conversions coming from brand search. The structural causes are known: over-segmented accounts that starve smart bidding of signal, targeting drift, and platforms optimizing toward cheap conversions because the CRM outcomes never flowed back. The Outshine chapter closed Aug 31; Verto took over search and LinkedIn Sep 1 with keep-the-lights-on continuity, which makes this the clean-slate moment to rebuild rather than patch.

The rebuild, per piece (SOW §3.2 + ratified decisions)

• Account audit + the Redis keyword portfolio — a one-time audit of the existing search accounts for structure and conversion opportunities, and one centralized keyword/job-title source of truth (single spreadsheet, director-level vs practitioner-level targets distinguished) so targeting stops living in tribal memory.
• The consolidated ad-group architecture (ratified Aug 27) — fewer, bigger campaigns replacing per-region fragmentation, restructured around smart-bidding signal thresholds with budget floors per line so the engines get optimization volume. Verto's structural bar: enough impression and conversion volume per campaign that smart bidding actually learns.
• Bidding re-aimed at revenue — conversion media across Google Ads and Microsoft Advertising (software category, competitor, and brand terms) optimized toward direct SQL and Opportunity generation rather than clicks: each CRM stage a discrete, value-weighted offline conversion, value-based bidding tied to Opportunity and Closed-Won as pipeline data accumulates (the loop Data & tracking builds).
• OCT finalization — offline conversion tracking closed out from the Outshine era and rebuilt on the new rail, so no learning is lost in the transition.
• New-format testing (Month 2) — the ad testing plan across LinkedIn Message Ads and Text Ads, with Fibbler measuring account-level ad engagement; emerging surfaces (G2, ChatGPT advertising) run as scoped experiments, not always-on channels. Testing follows the system rule: two variables per test, judged by concept.
• Scheduled Meeting Program (Month 3) — the CPL meeting program: meetings as the conversion event, priced and measured.
• Message match end to end — high-intent traffic lands on pages that match the ad (the LP system from the website bet); creative ships bespoke per campaign audience with an asset taxonomy at the ad level so side-by-side testing reads clean.

The four buckets (where ~80% of the budget goes)

Walked through with Shafiq Sep 1 — the performance budget concentrates into four buckets, with awareness running separately:

• Brand search, small and exact (~20%) — optimized to the exact term: a search for Redis Flex lands on the Flex page, the brand ad carries the AI narrative plus sitelinks routing to multiple conversion destinations. Spend controlled, bid strategically — never disrupt what someone actually searched for. Framing: this is building brand equity in the *new* products, not divesting from brand.
• Non-brand, in-market terms (~40%) — keywords that assume someone is in market to buy, from the portfolio-wide term review. Where the term is net-new (agent memory), go high on search impression share and own it — accepting different-looking efficiency metrics to buy the position early.
• Retargeting by journey stage — LinkedIn, Reddit, X, and every retargeting surface: the right offer in the right place based on where the account sits in the buyer's journey. Enterprise-interest accounts get an entirely different treatment than a two-second trial-page visit.
• Competitor terms, selectively — competitor keyword lists loaded through the new campaign types on every vendor: no-brainer on smaller competitors, strategic on the big ones (we don't out-bid Mongo head-on).

Two standing rules ride along. OSS-signup exclusion: done paying for open-source signups and downloads — build exclusions as accounts sniff around OSS. Content offers don't produce hand-raisers: ebooks and gated content feed the all-lead bottoms-up forecast, not the MQL line — which is exactly why the forecast must cover every lead type.

Awareness runs inside this project as a separate pure play — the section below — never graded on CPL.

The awareness pure play (absorbed Sep 4 — was Awareness Paid Media)

Sustained brand presence with proper brand-marketing KPIs, run while performance captures the demand it creates — Verto's paid growth associate (Stanislava) executes SOW §3.3 at ~75/25 awareness-to-lead-gen weighting, adjustable monthly. Measured on reach and account penetration, never CPL: awareness ran as leftover budget with lead-gen scoreboards before, which guaranteed it always looked like a failure.

• 01 · Connected TV — the brand in front of technical decision-makers at scale; the executive-reach rail.
• 02 · Paid social with brand KPIs — Reddit, X, YouTube, Meta + CarbonAds on the ICP segments; AI prospecting audiences from the firmographic database against closed-won seed data.
• 03 · Newsletters and podcasts — where builders and executives already spend attention; buying moves to BuySellAds/Carbon to cut direct-sponsorship overhead.
• 04 · OOH & experimental — AdQuick for short-term out-of-home; ChatGPT Ads as an early-mover experiment.
• The scale mechanism: 20–30 voices — our own people as the faces of the brand, equipped with greenscreens and HD kits, so awareness creative compounds without an agency production line.

Sequencing: Month 1 scope + KPI agreement and Phase 1 live (LinkedIn/X/Reddit) · Month 2 BuySellAds, podcasts, dev influencers, digital OOH · Month 3 measure lift, cut underperformers, expand on results.

Operating rules

Targeting locked to the ICP with exclusions enforced in every platform; the same three personas as email (builders, executives, customers); a documented strategy for every dollar by region and sub-region; partner co-funding extends the budget where the math works. Budget split: performance 65% / awareness 35%. First ship (memo): the campaign architecture doc.

Work to be done

• Accounts rebuilt on the consolidated structure — audit complete, keyword portfolio centralized, OCT finalized, and the new search + LinkedIn architectures live on the ad-group model with budget floors and ICP exclusions enforced.
• Bidding optimizing to pipeline — CRM-stage offline conversions flowing, value-based bidding on Opportunity/Closed-Won, and the paid scoreboard reading pipe-to-spend against the ≥1.8x bar.
• New formats proven or killed — Message/Text Ads and Fibbler measurement through the two-variable testing rail, G2/ChatGPT experiments read out, and the Scheduled Meeting Program delivering meetings at target CPL.
• The awareness KPI frame agreed before spend scales — reach, account penetration, and brand engagement on its own scoreboard, reported separately from lead-gen so neither corrupts the other.
• All four awareness channels live in sequence — Phase 1 (LinkedIn/X/Reddit) then Phase 2 (BuySellAds, podcasts, dev influencers, digital OOH, CTV), each placement carrying a benchmark and a kill rule.
• The voices program producing — 20–30 internal voices equipped and publishing on a cadence, feeding paid social and the content engine with authentic creative.

Current state

Transition executed on our terms: Outshine hard stop Aug 31, Verto live Sep 1, Reddit off, keep-the-lights-on plans for Google Search and LinkedIn handed over. The offboard checklist is closing with campaign-by-campaign keep/pause decisions; quarter-to-date spend is being reconciled for September/October.

Dependencies & risks

Learning-phase resets are the physics of a restructure — expect a soft first weeks and judge on the trend, not week two. The revenue loop depends on Data & tracking's conversion pipelines landing on schedule (day-30 bar: real values feeding bidding). The audience layer (segments, exclusions) shares the 6sense configuration timeline. [owner to complete]