LinkedIn Lead Generation for a Lahore Retail-Operations SaaS

The Client

The company is a Lahore-based B2B SaaS business selling retail-operations software: point of sale, inventory management, and shrinkage analytics for multi-store Pakistani retailers. Its customers are pharmacy chains, grocery and superstore groups, and fashion retailers running anywhere from ten to eighty outlets. ARR sat around PKR 240M, growing steadily on a classic sales-led motion: an account executive demo, a thirty-day pilot in two or three stores, then a staged rollout contract worth PKR 2.4M to 6M over the first year. Four AEs and two SDRs owned the pipeline.

Growth had come almost entirely from founder networks and referrals, plus brand search picking up whatever reputation generated. Paid acquisition had been tried exactly once: a LinkedIn experiment running about PKR 0.5M a month, built as a single campaign targeting “retail professionals in Pakistan,” mixing boosted founder posts with static “Book a demo” ads. It produced roughly sixteen demo bookings a month at PKR 31,500 each — and the demos were a problem in themselves. A large share of respondents were job-seekers, students, or owners of a single store who either churned during the pilot or were never a fit for a multi-location product. No-shows ran at 44%. The AEs had started quietly deprioritising LinkedIn-sourced meetings, and the head of sales had cut the budget to a token level.

The brief to WeProms was to decide the channel properly: either run LinkedIn lead generation with real discipline around fit and follow-through, or stop spending on it. We took the first option, and the constraint shaped everything: no budget increase would be justified until cost per attended, qualified demo came down.

The Problem

The audit surfaced six issues, all of them compounding:

• The audience was a category, not a customer list. One campaign covered every seniority, company size, and retail format in Pakistan. A director of retail ops at a forty-store pharmacy chain and a freshly graduated job-seeker were paying the same auction price for entirely different value.
• Cold traffic was asked to propose marriage. The only call to action was “Book a demo” — a high-commitment ask aimed at an audience that had never heard of the product, with no intermediate step to build trust or qualify interest.
• No exclusion discipline. The Pakistani LinkedIn member base skews heavily toward job-seeking and student traffic when targeting is broad. Nothing in the setup filtered it out, so a meaningful share of budget bought people who could never buy.
• The follow-up was a shared inbox. Lead gen form submissions arrived as notification emails. SDRs worked them when they got spare cycles, typically two to three days later, by which point whatever intent existed had cooled.
• The algorithm was optimising to the wrong event. LinkedIn could only see form fills — it had no signal for demos held, let alone SQLs. So bidding concentrated on whoever was cheapest to convert, which was precisely the job-seeker traffic poisoning the funnel.
• Nobody could connect spend to pipeline. There was no attribution from ad to demo to opportunity, so the only available conclusion was the blunt one the team had reached: “LinkedIn doesn’t work for us.”

The underlying failure was structural: the channel was being asked to produce qualified demos while being instrumented, targeted, and measured for raw form fills.

Phase 1 — Funnel Audit and Benchmark-Report Build (Weeks 1-2)

The first two weeks produced no ads and no creative. They produced the two assets everything else depended on: a real ICP and a lead magnet worth a business phone number.

Win-loss analysis from the CRM. We pulled two years of deals and segmented closed-won contracts by retail format, store count, and cycle length. Three patterns held: pharmacy chains with fifteen-plus stores closed fastest, because expiry-led shrinkage is a measurable, urgent cost; grocery and superstore groups carried the largest first-year contracts; fashion retailers closed slowest, with the longest pilot-to-rollout lag. Critically, 81% of closed-won revenue came from chains with ten or more stores — single-store and sub-ten-store retailers churned in pilot at rates that made them a cost, not a pipeline. The prior campaign had been spending a meaningful share of budget against exactly that segment.

Segment	Typical stores	Cycle	First-year value	Demo priority
Pharmacy chains	15–60	5–7 weeks	PKR 2.4M–4M	Highest
Grocery and superstore groups	20–80	8–12 weeks	PKR 3.5M–6M	High
Fashion retailers	10–40	12–16 weeks	PKR 2.4M–3.5M	Selective
Single-store owners	1–3	—	Churn trap	Excluded

Audience architecture. We rebuilt targeting per segment: job titles covering owners, directors, and heads of retail operations and IT, seniority floored at Manager with the weighting toward Director-plus, and explicit exclusions for entry-level titles and job-seeker-heavy groups. Each segment got its own campaign, so creative, bid, and budget decisions could differ where the buyers differed.

The lead magnet. Here the client held an asset most Pakistani SaaS companies have and never use: aggregated, anonymised operational data from roughly 1,400 storefronts on its platform. We commissioned the Pakistan Retail Operations and Shrinkage Benchmark — stock-out rates, shrinkage as a percentage of revenue, inventory accuracy, and dead-stock share, split by pharmacy, grocery, and fashion formats. It gave a retail operator something a product ad never could: a mirror showing how their chain compared. Gating it with a lead gen form turned that curiosity into identifiable pipeline. A quarterly refresh was planned from the start, making it a recurring asset rather than a one-off PDF.

Production discipline mattered as much as the idea. The benchmark drew only on aggregated, anonymised platform data with no chain identifiable in any cell, it was reviewed for sample-size honesty — formats with thin coverage were flagged rather than quietly averaged — and it was typeset to look like reference material a head of retail would forward internally, not like marketing collateral. That single design decision shaped who downloaded it: the report spread internally inside prospect chains, which manufactured the multi-stakeholder visibility a sixty-store purchase requires and that no single ad could buy.

Phase 2 — Funnel Architecture and Closed-Loop Tracking (Weeks 3-5)

With the audience and the offer defined, we built the plumbing that would let the channel learn toward revenue instead of form fills.

Qualification at the form. The lead gen form carried four qualifying questions alongside contact fields: store count band, retail format, current POS or spreadsheet usage, and intended decision timeline. Each answer wrote to a CRM field on lead creation. This did two jobs at once — it let SDRs triage before spending a call, and it gave the funnel a structure for scoring later.

Landing pages per segment. Report traffic landed on a short page per retail format, leading with that format’s benchmark numbers. A separate demo page carried an embedded calendar, a two-minute product walkthrough recorded for each vertical, and transparent pilot framing — pricing model, pilot length, rollout path — so that whoever booked a demo arrived pre-qualified on mechanics, not just curiosity.

The CRM integration. Form submissions stopped emailing an inbox. They routed into the CRM with segment, store band, and timeline attached, stamped with campaign and creative UTMs. A five-step nurture sequence carried downloaders who were not ready — each email unpacking one benchmark finding with practical commentary, no product pitch until step four.

Offline conversion sync — the core of the build. We defined three conversion events and pushed them back into LinkedIn from the CRM with values attached: Demo Held (the meeting actually happened), SQL (sales-accepted fit after discovery), and Closed-Won (staged first-year value). This is what changed the channel’s objective function. Once synced, LinkedIn’s bidding could start distinguishing the cheap job-seeker fill from the pharmacy-chain director — because only one of them ever produced a Demo Held event.

Phase 3 — Launch, Creative Testing, and Optimisation (Weeks 4-8)

Campaigns went live in week four across three vertical lead gen campaigns, one retargeting layer, and a low-budget demo-conversion campaign for warm audiences.

Creative built from the benchmark, not from features. The testing grid crossed three angles with three formats. The angles: benchmark stat cards (“Pakistani pharmacy chains lose an average of 2.6% of revenue to expiry-led shrinkage — where does your chain sit?”), a founder POV video on why shrinkage data is broken across multi-store retail, and a customer-proof carousel walking an anonymised chain’s pilot-to-rollout numbers. Formats were single image, short native video, and carousel. The stat cards generated the cheapest qualified leads; the founder video produced the highest demo rate among people who watched past the halfway point. Average CPC fell from about PKR 1,850 to PKR 1,480 across the phase as relevance improved — LinkedIn’s auction rewards ads the targeted audience actually engages with.

Bidding migration as events accumulated. Lead gen campaigns launched on cost-cap bidding against the form-fill event to establish volume. Once roughly thirty Demo Held events had synced, the warm retargeting campaign switched to a conversion objective optimising toward Demo Held — and this is where performance separated. The algorithm began selecting for the people who look like demo-holders, not the people who look like form-fillers. By week eight, cost per demo had fallen 26% from baseline while volume more than tripled.

Retargeting layers. Report downloaders who had not booked saw a sequenced message set: benchmark findings, then the walkthrough video, then a direct demo invitation with the pilot framing. Demo-page abandoners who had loaded the calendar but not booked received a single reminder with a vertical-specific proof point. Frequency caps kept retargeting spend from recycling against converters.

Metric	Before	End of Phase 3 (Week 8)
Demo bookings (monthly)	16	53 (3.3x)
Cost per demo	PKR 31,500	PKR 23,400
Demo no-show rate	44%	26%
Lead-to-demo rate	13%	21%

The no-show improvement told the real story: when the person booking actually manages retail operations at a multi-store chain, they show up. The previous no-show problem had been an audience problem wearing a scheduling costume.

Phase 4 — Sales Handoff and Compounding (Weeks 8-12)

The final phase hardened everything that happens after the click — where the in-house effort had quietly died.

Scoring and routing. We implemented lead scoring and sales handoff optimization on top of the form-answer fields: firmographic fit weighted store band and format against the Phase 1 ICP table, and intent weighted timeline, walkthrough-video engagement, and prior downloads. High-fit, high-intent leads routed directly to an AE with the qualifying answers attached so the first call started as a working conversation. Mid-fit leads entered the nurture track. Sub-ten-store single-store submissions were politely disqualified automatically with a self-serve resource — the SDR team stopped burning hours on them entirely.

A same-business-day contact SLA. Every routed lead received first human contact within the same business day, with talk tracks per vertical: expiry and stock-out math for pharmacies, dead-stock and inventory accuracy for grocery, inter-store transfer leakage for fashion. Speed mattered less than the in-house effort’s assumed — what mattered was calling with the lead’s own benchmark context in hand.

Monthly creative refresh from the report. As new quarters of benchmark data came in, stat-card creative refreshed with fresh numbers, which kept click-through rates from decaying and gave the nurture sequence a steady stream of genuinely new material. The lead magnet became a flywheel asset rather than a frozen PDF.

The pipeline dashboard. One closed-loop view ran spend to form submissions to attended demos to SQLs to pipeline, segmentable by vertical and creative. For the first time, the head of sales could see which rupees produced which opportunities, and budget reviews became arithmetic instead of argument.

By week twelve the channel was running at about PKR 1.33M a month — 2.6x the original budget, unlocked strictly by the efficiency gains — and producing sixty-seven demos a month at PKR 19,850 each, with demo-to-SQL at 44% and roughly PKR 71M in qualified pipeline added over the quarter, measured on staged first-year contract value.

Final Results

Metric	Before	After 90 days	Change
LinkedIn-sourced demo bookings	16/month	67/month	4.2x
Cost per demo	PKR 31,500	PKR 19,850	-37%
Lead-to-demo rate	13%	27%	+108%
Demo-to-SQL rate	31%	44%	+13 pts
Demo no-show rate	44%	19%	-25 pts
Qualified pipeline (quarter)	Not tracked	~PKR 71M	New visibility
Monthly channel spend	PKR 0.5M	PKR 1.33M	+2.6x, unlocked by efficiency

Cost per lead, for the record, rose about 30% across the engagement. That was the correct trade: the funnel was buying fewer cheap junk submissions and more expensive qualified ones. Cost per lead is a vanity metric in a sales-led SaaS; cost per attended, qualified demo is the scoreboard.

These figures are illustrative outcome ranges reflecting patterns WeProms sees in Pakistani B2B SaaS demand generation, not an audited result for a named company. They give a growth team a realistic shape for what a lead-magnet-and-closed-loop LinkedIn program can produce.

What Made This Work

1. The lead magnet came from data only the client had. Anyone can gate an ebook. A benchmark built from 1,400 Pakistani storefronts gave retail operators a genuinely scarce answer — how do I compare — and that is what converts a cold audience into identifiable, triagable pipeline.
2. Qualification happened at the form, not on a call. Four questions filtered non-buyers before SDR time was spent, doubled the lead-to-demo rate, and doubled again the value of every hour the sales team invested in the channel.
3. The algorithm was finally optimising toward revenue. Syncing Demo Held and SQL back as offline conversions was the mechanism that cut cost per demo — bidding stopped rewarding whoever was cheapest to convert a form and started rewarding whoever looked like a future customer.
4. Exclusions did quiet, expensive work. Filtering job-seeker and entry-level traffic, and excluding the sub-ten-store churn trap the CRM had already identified, removed spend that could never produce pipeline no matter how cheap it looked.
5. Budget followed evidence, in order. Spend scaled 2.6x only after cost per demo had fallen — the efficiency unlocked the scale, rather than scale being bet on in hope of efficiency.

What Teams Can Apply

For Pakistani B2B SaaS companies running a sales-led motion on LinkedIn:

1. Audit your CRM before your ad account. Closed-won data tells you who actually buys and, just as usefully, who churns in pilot. Exclude the churn segment ruthlessly — it is the most expensive cheap traffic you own.
2. Give cold traffic a stepping stone, not a demo form. A benchmark, calculator, or diagnostic rooted in local data earns the hand-raise. Then let nurture and retargeting carry the warm ones to the demo.
3. Sync offline conversions or accept that you are bidding blind. Until LinkedIn can see which fills become demos and SQLs, its bidding will optimise toward the cheapest humans on the platform. This integration is the single highest-leverage technical step in the whole playbook.
4. Qualify at the form with questions you will actually use. Store count, format, current system, timeline — each answer should drive routing, scoring, or talk tracks. If a question changes nothing downstream, delete it.
5. Judge the channel on cost per qualified demo and pipeline, and say the cost-per-lead number out loud when it rises. A rising CPL alongside a falling cost per demo usually means the funnel has started buying better humans.

WeProms Digital has run this lead-magnet-and-closed-loop pattern across Pakistani B2B software companies in retail tech, HR and payroll, ERP, and logistics. The benchmarks and verticals change; the sequence — ICP from CRM data, a data-backed offer, qualification at the form, offline conversion feedback, disciplined handoff — does not. The same structure supports our broader work in digital marketing for SaaS selling from Lahore and Karachi into domestic and regional accounts.