Performance Max is a Google campaign type that spends across Search, Display, YouTube, Discover, Gmail, and Maps from one budget. It is optional. For most Nepal SMEs it is later, not day one. Arjan KC treats it as a technique on SEM, not a default package. Term: Performance Max. Setup encyclopedia: PMax guide.

→ Google Ads Nepal · SEM as one plan · Account ownership

Published Search results, not PMax

These are Search rebuilds:

A PMax percentage is not published on this page. CPC context: Google Ads cost Nepal.

Why Google Ads fail without structure in Nepal

Image: Arjan KC media library.

Go / no-go

Start PMax when… Wait when…
Purchase or qualified lead fires in GA4/Ads Conversion is “WhatsApp click” or missing
Dedicated landing, not / Homepage is the only URL
Enough conversions/month for value bidding A handful of calls logged in a notebook
Product and category pages live on your own site Catalog lives only on Instagram
Brand Search exists separately PMax would eat the company name

COD and inbox closes mean last-click ROAS is incomplete. Smart bidding that cannot see the sale will chase the proxy you gave it — often the cheapest empty click.

Google Ads failure patterns

Image: Arjan KC media library.

Nepal order of operations

  1. Own the account — ownership.
  2. Query map + negatives — audit · negative keywords.
  3. One landing, one CTA — form, call, or documented chat.
  4. Search until you trust the conversion.
  5. Then PMax (or Performance Max + Search), with brand exclusions.

Festival weeks (Dashain) are a bad moment to switch the account to fully automated if tracking still lies. Paid seasonality: festival campaign SLA · seller checklist.

What PMax needs from your account

Smart bidding learns from conversions. Google’s own guidance asks for a steady monthly count of the conversion you want it to optimise; if you cannot say how many qualified leads or purchases your account recorded last month, you are not ready, whatever the number is. Three further requirements:

  • A conversion that means money. “WhatsApp click” or “thank-you page viewed” trains the system to find clickers, not buyers. Import confirmed sales or qualified leads (offline conversions) and make that the primary conversion.
  • A landing page per asset group. Education consultancies send IELTS and Australia-study queries to different pages; a clinic sends dental and dermatology queries to different pages. PMax will use the homepage if you let it.
  • Brand excluded. Add brand terms as account-level negatives or keep a separate brand Search campaign. Otherwise PMax reports your own name searches as its wins.

Asset groups for a Nepal account

Write assets the way the landing page speaks: service, area, NPR price or starting price, and one call to action. Mix Nepali and English headlines only if the landing page does the same. Use your own photos and short vertical video; stock imagery of foreign offices lowers trust and the system will still spend on it. Audience signals should be your customer list and converters, not interest categories Google guesses.

Read the report honestly

PMax reports fewer details than Search. Check three things monthly: the search terms insights for queries that belong in Search instead, the placement report for Display and YouTube spend on irrelevant apps, and the share of conversions that are brand. If brand dominates, the campaign is cannibalising, not growing. Compare cost per qualified lead against the Search campaign over the same weeks. The published Search rebuilds on this site (−42% CPA, −40% CPL) were achieved with structure and negatives, not automation; that is the baseline PMax has to beat.

A test you can afford to lose

Run PMax for six weeks (Google’s own minimum for judging a new campaign, see below) with a capped daily budget alongside the existing Search campaign, not instead of it. Agree the success measure before launch: cost per confirmed lead or order, imported from your CRM or sheet. Industry CPC bands from the benchmark report give you the ceiling to compare against. If PMax cannot match Search on that number by week six, pause it and keep the learnings for next quarter. Write the result down either way; the next person who proposes “just switch everything to PMax” should have to argue with your own account data, not with this article.

Setting up Performance Max for a Nepali online store

Everything below assumes the go/no-go table says go, and that you sell from your own website to buyers in Nepal. The generic walkthrough (asset specs, tROAS maths, account structure) is in the PMax campaigns guide; this section covers what is different for a Nepali store.

First check: Shopping ads cannot target Nepal

Most PMax guides describe the product-led version, where Google pulls product listings from a Merchant Center feed. That only works in a target country where Google supports Shopping ads and free listings. As of 13 October 2026, Nepal is not on that list. Google’s supported countries and currencies table and its country-not-supported help page both omit Nepal, and Nepali is not one of the supported feed languages. India, the UAE, the UK, the US and Australia are on the list.

Google has edited this list more than once in 2026, so open both pages again before you build anything. Until Nepal appears:

  • A feed targeting Nepal will not put your products into Shopping ads or free listings in Nepal. Do not plan budget for product-listing ads here.
  • PMax for a Nepal-only store runs without a feed: asset groups of text, images and video that point at your category and product pages, bidding towards a purchase. Google publishes separate optimisation tips for online sales without Shopping.
  • Merchant Center is not a launch blocker for a Nepal-only store. Feed hygiene in the Merchant Center Nepal guide matters if you export or if Nepal is added later.

The exception: Nepali exporters

A business in Nepal that ships abroad (tea, pashmina, handicrafts) can run feed-based PMax into supported countries. Google’s cross-border rules ask for product data in a supported language, prices in the target country’s currency (or Google’s currency conversion), shipping settings for that country, and landing pages and checkout in the same language as the feed. In practice: an English feed, a checkout that takes international cards, delivery times you can actually meet, and a separate campaign per target country so budgets and returns do not blur. Google Ads billing from Nepal is still in USD; see dollar card billing.

Prerequisites before the first rupee

  1. Company-owned accounts. Google Ads, GA4 and Tag Manager under the business’s own login (ownership).
  2. A purchase conversion with value and an order ID. Google lets you create PMax without conversion tracking but warns that spend may be limited. Send the order value in NPR and a unique order ID with every purchase; the ID is what lets you correct the count later.
  3. Every payment path reports, honestly.
    • eSewa, Khalti, Fonepay or card: fire the purchase on your own order-confirmation page after the gateway returns success, not on the button that sends the shopper to the wallet. Otherwise everyone who gives up at the wallet login counts as a buyer. Gateway setup: payment gateways in Nepal.
    • Cash on delivery: an order placed is not a sale. Record it with its order ID, then retract or restate refused and returned orders. Google says adjustments reach automated bidding only if uploaded within 7 days of the conversion (reporting accepts them up to 54 days), so a weekly upload from the delivery sheet is the minimum rhythm.
    • WhatsApp closes: import them as offline conversions against the original click, or leave them out. Never make “WhatsApp click” the primary goal.
  4. Recipes: tags and triggers for all of the above are in GA4 and GTM conversion tracking recipes.

Budget and learning period: what Google actually says

Google does not publish a fixed minimum conversion count for PMax. It does publish two rules worth following:

Do the arithmetic before launch. Illustrative: if Search confirms orders at NPR 1,500 each, three times that is NPR 4,500 a day, roughly NPR 135,000 a month in media. That is close to the “e-commerce scale” tier (NPR 150,000+/month) in the benchmark report. If your budget looks more like the report’s lead-gen test tier (NPR 20,000–40,000/month), PMax will not have enough room to learn; keep the money in Search.

Asset group structure

Build one asset group per product category that has its own landing page and margin (for example women’s kurtas, men’s shoes, festival gift boxes), not one per product and not one for the whole shop. Set the final URL to the category page. If you only want traffic on the pages you chose, switch off Final URL expansion, which otherwise lets Google swap in other pages from your site.

For online sales without Shopping, Google’s tips ask for at least 15 headlines, 5 descriptions, 7 images (3 landscape, 3 square, 1 portrait) and at least one video per asset group. Upload your own short video; Google notes this matters if you want to limit spend on auto-generated videos. Nepal specifics:

  • NPR prices that match the landing page, and delivery areas and times you meet (“Kathmandu Valley next day” only if it is true).
  • Payment options shoppers look for: eSewa, Khalti, card, COD.
  • Your own product photos, not supplier or marketplace images.
  • Asset language that matches the landing page.

Add search themes for the category terms people in Nepal actually type, including Romanised Nepali if your Search terms report shows it. Google now allows up to 50 search themes per asset group, and they respect brand exclusions and account-level negatives.

Audience signals

Signals are hints, not targeting; PMax can still show ads to people outside them. Google recommends remarketing lists, custom segments and Customer Match lists. For a Nepali store that means site visitors and purchasers collected by the Google tag, a list of past customers who agreed to hear from you (if your account is eligible for Customer Match), and a custom segment built from your category terms and comparable shop URLs. Broad interest categories on their own add little. More on audiences: remarketing in Nepal.

Brand exclusions and negative keywords

  • Keep brand in Search. Google says a Search campaign with an exact match keyword identical to the query takes priority over PMax, unless that Search campaign is budget-limited.
  • Add brand exclusions to the PMax campaign so it cannot claim your own name searches.
  • Add negatives. PMax supports campaign-level negative keywords and negative keyword lists, plus account-level negatives. They apply to Search and Shopping inventory only. A Nepali store usually starts with job and vacancy terms, “free”, “second hand”, “wholesale” if you do not sell wholesale, and “how to make” terms. Starting list: negative keywords master list.
  • For Display and YouTube, use account-level placement exclusions and content suitability settings, because keyword negatives do not reach that inventory (search terms report help).

Reading results after week six

Run the three monthly checks in “Read the report honestly” above, using these reports:

  • Channel performance (Insights and reports → Channel performance) shows cost and conversions by channel (channel report help). If YouTube and Display take spend while almost all purchases come from Search, PMax is mostly expensive Search.
  • Search terms and landing pages for Performance Max lists the queries, the landing page and the ad format for each (search terms report help). Add irrelevant terms as negatives; move proven high-intent terms into Search as exact match.
  • Asset reporting: Google suggests allowing three to four weeks before replacing “Low” assets, not daily swaps.

Judge PMax on confirmed orders after COD retractions, compared with Search over the same weeks. Platform-reported conversions before retractions flatter it.

Launch checklist

  • Merchant Center country list checked today; Nepal-targeted campaign planned without a feed, or a feed only for supported export countries
  • Google Ads, GA4 and Tag Manager owned by the business
  • Purchase fires on the confirmation page after wallet success, with NPR value and order ID
  • Weekly COD retraction upload, inside Google’s 7-day bidding window
  • WhatsApp closes imported as offline conversions, or excluded from goals
  • Brand Search campaign on exact match; brand exclusions on PMax
  • Campaign-level negatives or a negative keyword list attached
  • One asset group per category, final URL on the category page, Final URL expansion decided
  • 15 headlines, 5 descriptions, 7 images and your own video per asset group
  • Signals: purchasers, visitors, Customer Match (if eligible), one custom segment
  • Daily budget at least three times cost per order, held for six weeks without major edits
  • Success measure (cost per confirmed order) written down before launch

Next: PMax setup guide · Merchant Center Nepal · PPC Nepal · hire Search Performance.