Performance Marketing in Nepal: Paying for Outcomes, Not Impressions (2026)
Performance marketing means every rupee of spend, and ideally every rupee of agency fee, is judged against an outcome the business actually values: a qualified lead, a paid order, a booked visit. In Nepal that sounds simple, but most “performance” campaigns are still judged on clicks and messages because nobody defined what a result is, or tracked what happened after the WhatsApp chat started.
This guide covers how to set that definition, how to set targets from your margins, which deal structures work, and where performance marketing breaks down.
Why “performance” often fails in Nepal
Three local realities make outcome-based marketing harder than it looks in global playbooks:
- Sales close offline. Many enquiries move from an ad to WhatsApp, Viber, or a phone call, and the sale happens days later in person or on cash on delivery. If nobody records that, the ad platform optimises for chat starts, not buyers.
- Cash on delivery hides failed orders. An order placed is not an order paid. Return-to-origin and refused parcels mean “purchases” in Meta Ads Manager can overstate real revenue.
- Reply speed varies. A campaign can deliver good leads that go cold because the sales team replied the next afternoon. That is a sales problem, but it shows up as a “bad campaign”.
Performance marketing works when these gaps are closed first. Otherwise you are paying for optimisation against the wrong signal.
Step 1: Define one qualified outcome
Write a one-line definition that both the business and the marketer sign off on. Examples:
- Lead-based service: “A lead with a valid Nepali mobile number, in our service area, asking about a service we offer, that our team reached within 24 hours.”
- E-commerce with COD: “An order delivered and paid, net of returns.”
- Clinic or school: “A booked appointment or counselling session that the person attended.”
Then decide where that outcome is recorded: a CRM, a Google Sheet, or your order system. If it is not recorded anywhere, it cannot be optimised or paid on. Related: UTM naming and offline conversions for WhatsApp and COD.
Step 2: Set targets from margin, not from benchmarks
Benchmarks from other countries, or even other Nepali businesses, rarely match your prices and margins. Work backwards instead.
Hypothetical example: a Kathmandu business sells a service at NPR 20,000 with a 40% gross margin, so each sale leaves NPR 8,000 before marketing. If one in four qualified leads becomes a customer:
| Measure | Calculation | Result |
|---|---|---|
| Gross profit per sale | 20,000 × 40% | NPR 8,000 |
| Break-even cost per acquisition | Gross profit | NPR 8,000 |
| Break-even cost per qualified lead | 8,000 × (1 ÷ 4) | NPR 2,000 |
| Target CPL leaving profit | Say 50% of break-even | NPR 1,000 |
For e-commerce, the same logic gives a break-even ROAS of 1 ÷ gross margin. At a 40% margin that is 2.5; anything below loses money on the first order unless repeat purchases make up the difference. The marketing calculators hub has the formulas.
Write these numbers down before launching. Arguments about whether a campaign is “working” end quickly when the target was agreed in advance.
Step 3: Make the outcome visible to the ad platforms
Platforms optimise toward the signal they receive. To feed them the real outcome:
- Track the first step reliably. Form submits, calls from ads, and WhatsApp clicks should fire clean conversion events through Google Tag Manager. See GA4 and GTM conversion recipes.
- Capture an identifier. Keep the click ID (gclid, fbclid) or UTM source alongside the phone number in your CRM or sheet.
- Send the qualified result back. Upload offline conversions to Google Ads, and use the Conversions API for Meta when the sale is confirmed. Enhanced conversions help match records when click IDs are lost.
- Respect consent. Configure consent mode and only upload customer data you collected with permission.
Once qualified outcomes flow back, bid strategies such as target CPA or target ROAS have something meaningful to work with. Before that, they optimise toward cheap, low-intent actions.
Step 4: Choose a deal structure that both sides can live with
Pure commission sounds attractive to business owners, but the marketer does not control price, stock, delivery, or how fast your team replies. Common structures that hold up:
| Model | How it works | Works best when |
|---|---|---|
| Fixed fee + performance bonus | Base retainer, bonus above an agreed qualified-lead or revenue target | Tracking is in place and targets are agreed |
| Pay per qualified lead | Fixed rupee amount per lead meeting the written definition | Lead definition is strict and disputes are rare |
| Revenue share | Percentage of tracked revenue | E-commerce with clean, paid-order tracking |
| Fixed fee only | Retainer judged on agreed KPIs | Tracking is still being built, or volumes are small |
Whatever the model, agree in writing on: the outcome definition, the source of truth (CRM or order system, not the ad platform), the attribution window, and who pays ad spend. See pay per lead for the trade-offs of lead-based pricing.
Step 5: Run a weekly performance review
Keep it to one page:
- Spend, qualified outcomes, and cost per qualified outcome, by channel.
- Lead-to-sale rate from the CRM, and median reply time.
- Top three search terms or audiences by qualified outcomes, and the bottom three by waste.
- One decision for the coming week: what to scale, what to pause, what to test.
If the review starts with impressions or reach, it is a brand report, not a performance review.
Where performance marketing breaks down
- Very small budgets. Bid strategies need enough conversions to learn. With only a handful of results a month, manual control and tight targeting usually beat automation.
- Long or relationship-driven sales. Education counselling, real estate, and B2B deals can take weeks. Optimise toward a mid-funnel step you can record, such as an attended counselling session, and review revenue monthly.
- Brand-new demand. If nobody searches for your product yet, pure performance channels have little to capture. You will need awareness work, and it should be judged differently.
- Over-crediting. Platforms each claim the same sale. Use your own records, and occasionally test incrementality by pausing a channel in one area or period.
Checklist
- One written definition of a qualified lead or sale, agreed by sales and marketing.
- A target cost per outcome worked out from your margin and close rate.
- Conversions sent to the ad platforms, with offline outcomes imported where possible.
- A deal structure with a base fee, a performance element, and clear rules for disputed leads.
- A weekly review that compares platform numbers with your own CRM or order records.
- A small awareness budget, judged separately, if demand for your product is still new.
Common mistakes
- Paying a bonus on platform-reported conversions instead of CRM-confirmed results.
- Counting every WhatsApp chat start as a lead.
- Changing targets mid-month when results dip, instead of reviewing the full period.
- Scaling spend before reply speed and follow-up can handle more leads.
- Ignoring returns and refused COD parcels when calculating ROAS.
Key takeaways
- Define one qualified outcome and record it somewhere other than the ad platform.
- Set CPL, CPA, or ROAS targets from your own margins and close rates.
- Feed qualified results back to Google and Meta so bidding optimises for buyers.
- Use a fixed fee plus bonus when the marketer cannot control the whole sale.
- Review weekly on one page, with one decision.
How this fits the stack
The service itself is described on Performance Marketing Nepal, and the background behind it on the performance marketing expert page. For channel choice, read Google Ads vs Meta Ads for Nepali businesses. For lead-based businesses, lead generation in Nepal covers channels and follow-up.
Related dictionary terms
Next step
If your tracking stops at “message sent”, start there. Scope an audit through hire packages or contact.
