YouTube Demand-Gen CAC Reduction for a SaaS Startup

The Client

A Karachi-based B2B SaaS startup selling a field-sales productivity platform — call reporting, route planning, follow-up automation, and pipeline visibility — to Pakistani SMEs and small Gulf distributors. The company sat at roughly 420 paying accounts, PKR 22 million in monthly recurring revenue, an average contract value near PKR 52,000 per year, and a gross margin around 78%. Retention was healthy and the product held its own in head-to-head evaluations.

The economics of growth were the problem. The company was sales-led: an inside-sales team booked demos, closed accounts, and handed off to a small onboarding function. Acquisition ran almost entirely through Google Search (branded defense plus a broad bucket of category keywords), with some Meta lead-gen and LinkedIn outreach filling the gaps. Over twelve months, three well-funded competitors had entered the category and bid up the same keywords. Search CPCs in the bucket had inflated roughly 35%, blended CAC had climbed from PKR 27,000 to PKR 38,400, and payback had stretched past eleven months. The board had set a clear line: bring blended CAC back under PKR 30,000 without throttling pipeline.

The deeper issue was structural. The company had no demand-gen layer at all. Every lead was bottom-funnel — someone actively searching for a sales tool that month. There was no work being done to make the next quarter’s buyers aware that the category, and this product, existed. They were competing for the same small pool of high-intent searches while a much larger pool of eventual buyers went unreached.

The Problem

The diagnostic surfaced six issues, only one of which was a media-mix question:

1. Blended CAC at PKR 38,400, up 42% year-on-year. Payback had passed eleven months against a target under nine. The number was getting worse each month, not better.
2. Budget concentration. Roughly 84% of paid spend sat in Google Search, increasingly in broad-match buckets where CPCs were rising fastest. Diminishing returns were visible in the curve.
3. No demand-gen layer. Brand-search volume had been flat for three quarters. The team was paying for intent it had not helped create, in a market where competitors were outbidding it for that same intent.
4. Broken measurement. Enhanced conversions were off, there were no view-through conversion windows, and closed-won data from the CRM was not being imported back into the ad platforms. The reported CAC understated any channel that assisted rather than clicked last — which is exactly how a demand-gen channel behaves.
5. No creative system. The account had two static search ads and zero video assets. There was nothing to test, and no framework for deciding what to scale or cut.
6. Eroding lead quality. As the team broadened keywords to keep volume up, the sales cycle had lengthened to 7-9 weeks and the close rate on inbound demos had slipped. Chasing lead volume at the top was degrading economics at the bottom.

Phase 1 — Measurement and Baseline Rebuild (Weeks 1-2)

The engagement opened with measurement rather than media, because the team could not afford to judge a demand-gen channel against a tracking setup that penalized it.

Conversion tracking rebuild. We implemented enhanced conversions in Google Ads, configured the qualified-demo and closed-won events in GA4, and built a CRM offline conversion import that pushed HubSpot stage changes — demo booked, opportunity created, closed-won — back to the ad platforms against both click and view interactions. This single change reattributed a meaningful slice of pipeline that had been silently landing on “direct” and “organic”.

Defining CAC in tiers. A single blended number hides the real story. We split CAC into three tiers — lead CAC, qualified-opportunity CAC, and closed-won CAC — and baselined the blended (closed-won) figure at PKR 38,400. The team’s internal reporting had been tracking lead CAC, which looked healthier and masked the deterioration.

Enabling view-through. We turned on 30-day view-through conversion windows so YouTube’s assisted impact could be measured. This is the technical precondition for any honest demand-gen evaluation; without it, a channel that shows an ad to a buyer who later searches and converts gets zero credit and looks unprofitable.

Creative inventory audit. The account held two search ads and no video. We documented the gap and queued creative production for Phase 2. The baseline picture is captured below.

Tracking signal	Before	After (Week 2)
Enhanced conversions	Off	On
View-through window	None	30-day
CRM offline import	None	Demo + closed
CAC tier reported	Lead only	3 tiers
Video creative assets	0	In production

Phase 2 — Creative System and Audience Build (Weeks 3-6)

With measurement honest, we built the engine that would actually move the number. In demand-gen, creative is the primary lever — more than targeting, more than bidding — so most of this phase was production and structure.

Four creative pods. We produced four pods mapped to the buyer journey rather than four variations of the same ad:

• Problem-aware — a 25-second spot naming the pain (“your field reps forget the follow-up the moment they leave the meeting”), cut for in-stream skippable placement.
• Solution demo — a 30-second screen-capture showing the two workflows that saved reps time, aimed at buyers already evaluating.
• Social proof — a short testimonial from a Lahore distributor (an illustrative, anonymized account) framed around measurable rep productivity gain.
• Shorts hook — a 15-second vertical hook built for YouTube Shorts and Demand Gen discovery placement.

Audience layering. We built audiences rather than buying default demographics: lookalikes seeded from the top decile of accounts by expansion revenue, custom-intent segments built from the category keywords competitors were bidding on, customer-match suppression to exclude current accounts and recently churned ones, and life-event and job-change signals to reach newly-promoted sales managers. Each audience was paired with the creative pod most likely to land for it — the problem-aware pod ran to cold custom-intent and lookalikes, the demo pod to retargeting and warm custom-intent, the social-proof pod to mid-funnel audiences who had visited pricing but not booked, and the Shorts hook to broad affinity as a cheap reach layer. Pairing audience to creative, rather than running every ad to every audience, is what kept cost-per-qualified-demo interpretable in the weekly scoring.

A deliberate constraint shaped the audience design: suppression. Current accounts, recently churned accounts, and existing pipeline contacts were suppressed across all paid media, not just YouTube. Demand-gen spend that reaches people already in your CRM is spend that cannot show return, and it inflates apparent reach while hiding the real cost of net-new demand. The detail of this creative-and-targeting system is covered in our YouTube ads strategy and management scope.

Placement mix. We set up skippable in-stream and non-skippable bumpers for the problem-aware and demo pods, Demand Gen video and image placements for retargeting and lookalikes, and a small YouTube Shorts pilot for the hook creative. The conversion objective across all of it was qualified-demo booked, not impression volume.

Phase 3 — Launch, Test, and Reallocate (Weeks 4-10)

Phase 3 is where spend moved and the CAC curve started to bend.

Sequenced reallocation. Rather than layering YouTube on top of an unchanged search budget, we shifted roughly 30% of paid budget out of the worst-performing broad-match search campaigns and into the YouTube demand-gen layer, while leaving exact-match brand defense untouched. The point was not to abandon search — it was to stop overpaying for saturated intent and use the freed budget to manufacture next quarter’s intent.

Weekly creative scoring. Every pod was scored weekly on cost-per-qualified-demo (click and view-through) and view-through rate. The bottom quartile was cut each week and replaced with a fresh hook; the top performers were fed more budget. Over six weeks the cost-per-view fell from roughly PKR 7.10 to PKR 4.20 as bidding optimized and creative quality rose.

The priming effect. Midway through the phase we noticed search conversion rates on brand and category terms climbing — up roughly 19% on the primed audience versus a geo holdout that had not seen YouTube. This is the real mechanism of demand-gen for a sales-led SaaS: it does not replace search, it makes search cheaper by warming the buyer before they ever type a query.

Second reallocation. With the priming signal confirmed, we moved another 12% out of the weakest search campaigns into the two winning creative pods. The shape of the budget had now meaningfully changed — search was still the largest line, but demand-gen was doing real work upstream.

Creative pod performance at end of Phase 3:

Pod	Placement	Cost-per-view	Cost-per-qualified-demo	Decision
Problem-aware	In-stream	PKR 5.10	PKR 3,900	Scale
Solution demo	In-stream	PKR 4.80	PKR 2,650	Scale
Social proof	Demand Gen	PKR 3.90	PKR 3,100	Scale
Shorts hook	Shorts	PKR 2.40	PKR 6,800	Refresh

Phase 4 — Scale and Compound Efficiency (Weeks 8-24)

The final phase ran across two quarters and turned a working test into a durable efficiency gain.

Scaling winners. The three winning pods were scaled to roughly PKR 2.4 million a month in YouTube spend. Scaling demand-gen creative is where most teams fail, because creative fatigues fast — we measured a half-life of roughly 18 days on a winning pod before view-through rate decayed. We held the line with a biweekly creative refresh, rotating new hooks into the proven structures rather than starting over.

Modeling assisted impact. We tracked YouTube-assisted closed-won through a blend of view-through conversions and a data-driven attribution model, reconciled monthly against the CRM. YouTube-attributed closed-won CAC came in at roughly PKR 21,800, against PKR 41,000 for search — not because YouTube was cheaper per lead in isolation, but because it was generating net-new logos that search alone would not have reached at any price.

Optimizing to opportunity, not lead. We shifted the campaign optimization target from lead CAC to qualified-opportunity CAC. This cut junk demo bookings, freed inside-sales capacity, and shortened the average sales cycle by roughly ten days. Tracking this honestly is the discipline behind our CAC and payback modeling work.

Incrementality check. A quarterly geo holdout confirmed YouTube was driving net-new logos rather than cannibalizing search. We split comparable regions into a treated group (which saw YouTube spend) and a holdout group (which saw none), then compared closed-won logos per region over the quarter while holding search spend flat across both. The treated regions produced roughly 19% more net-new logos than the holdout, with no compensating search lift in the holdout — meaning the demand-gen channel was genuinely additive and not simply reattributing demand that would have arrived through search anyway. This is the test that distinguishes a working demand-gen layer from an expensive reporting artifact, and it is the reason the team could defend the reallocated budget internally.

Avoiding the vanity metrics. Throughout the engagement we deliberately de-prioritized view count, reach, and impression share as optimization targets. A demand-gen campaign can hit any of those numbers cheaply and still produce no pipeline. The only metrics that moved budget decisions were cost-per-qualified-demo, view-through-assisted closed-won, and the incremental logo lift from the geo test. Keeping the dashboard narrow forced every decision through the lens of pipeline economics.

Final Results Over Two Quarters

Metric	Before	After	Change
Blended CAC (closed-won)	PKR 38,400	PKR 26,500	-31%
Paid-attributed new logos (quarter)	Baseline	+44%	Net-new
YouTube-attributed closed-won CAC	n/a	PKR 21,800	vs 41k srch
Search conv. rate (primed audience)	Baseline	+19%	Assisted
Qualified opportunities	Baseline	+38%	Compounding
LTV:CAC	2.6x	3.8x	+1.2x
Cost-per-view	~PKR 7.10	~PKR 4.20	-41%

Every line traces to a phase: the CAC drop to the reallocation and priming effect in Phase 3, the YouTube-attributed CAC to the view-through modeling in Phases 1 and 4, the opportunity lift to the opportunity-level optimization, and the LTV:CAC expansion to CAC falling while retention held. The shape — demand-gen priming cheaper search, view-through revealing the real contribution, creative as the scaling lever — is the realistic pattern for a sales-led SaaS fighting CAC inflation in a competitive category.

What Made This Work

Four factors drove the result:

1. Measurement was fixed before media moved. Without enhanced conversions, CRM import, and view-through windows, YouTube would have looked unprofitable and the engagement would have ended in week six. The first two weeks of tracking work were the highest-ROI work in the entire engagement.
2. Creative was treated as a system, not an asset. Weekly scoring, a cut-and-refresh cadence, and four pods mapped to the buyer journey beat any single “great” ad. Demand-gen lives or dies on creative volume and refresh discipline; the 18-day half-life is the number to plan around.
3. Demand-gen fed search rather than replacing it. The 19% search conversion lift on the primed audience was the mechanism that actually moved blended CAC. The goal was never to leave search — it was to stop overpaying for saturated intent and manufacture cheaper intent upstream.
4. Optimization target moved from lead to opportunity. Chasing lead CAC had been quietly degrading the funnel. Once the team optimized to qualified-opportunity CAC, junk demo volume fell, sales capacity freed, and the cycle shortened. The headline metric you optimize to shapes everything downstream.

What Teams Can Apply

For a Pakistani B2B SaaS team watching CAC climb in a competitive category:

1. Fix view-through and CRM offline import before judging any demand-gen channel. If your tracking only credits last-click, a demand-gen channel will always look unprofitable and you will kill it before it compounds. This is a two-week prerequisite, not an afterthought.
2. Treat creative as the primary lever. More than targeting or bidding, creative quality and refresh cadence determine demand-gen outcomes. Plan for four to six pods at launch and a roughly three-week refresh rhythm on anything that scales.
3. Do not replace search — feed it. The cheapest way to lower search CAC is often to warm the audience first with demand-gen. Measure the priming effect with a geo holdout rather than trusting platform attribution.
4. Optimize to opportunity CAC, not lead CAC. Lead CAC rewards volume; opportunity CAC rewards pipeline quality. The shift frees sales capacity and shortens the cycle, and it usually moves the blended number more than any bid change would.

This illustrative engagement reflects the patterns WeProms Digital sees across Pakistani B2B SaaS companies fighting acquisition-cost inflation — the creative angles, audience builds, and reallocation ratios shift with each buyer profile, but the measurement-first, creative-system, feed-the-search sequence stays consistent. The broader environment of rising acquisition costs for tech startups in Pakistan is exactly the pressure this kind of demand-gen layer is built to relieve.