Advertising15 minSeptember 5, 2026
YouTube ads for services in Tashkent: when to test and what budget
Robert MirovUpdated Sep 5, 2026
People in Uzbekistan watch a lot of YouTube. That does not mean every clinic and every cleaning service should turn on a video campaign on Monday. This article is for a service owner in Tashkent: when video as a channel makes sense, which format to pick, what budget for a test, and why “drop a clip into Performance Max” is not the same as a controlled test. No “go viral” promises and no price list for views. UZNEO treats YouTube as a layer after working Search, or as a separate test with a clear goal — not “reach for the sake of reach”.
01Who YouTube fits — and who should wait
Video works when the service is visible: renovation, salon, kitchen before/after, auto service, dentistry (careful with policies), training. In 6–10 seconds a person should understand what it is and the next step (call, WhatsApp, form).
Too early if:
• there is no landing and no reply SLA — see when not to launch ads;
• there is no clip longer than 10 seconds that shows the actual service (not stock);
• Search still cannot count a lead;
• the test budget is smaller than you need for several hundred meaningful views with a chance to click.
YouTube is weaker at “urgent visit today” than Search for “master nearby”. For urgent demand, start with Search.
| Before | After | |
|---|---|---|
| “Everyone watches YouTube — let’s turn it on” | There is an offer on video and a way to send a lead | |
| Stock with no address and no language | Your own location, RU/UZ as clients speak | |
| Video instead of Search | Search for demand, video for warming up |
02Formats in plain words
In-Stream (TrueView) — a clip before/during a video, often with a skip button after 5 seconds. You pay for views under Google’s rules (usually when they watch past a threshold or click — check current terms in the account). The job: get the offer across before the skip.
In-Feed — a preview in the YouTube feed/search. Closer to “they chose it themselves”. You need a strong first frame and title.
Bumper — up to 6 seconds, no skip. For a brand reminder, not for a complex service with a price list.
Demand Gen / “Discovery” and video inside Performance Max mix placements. For a first service test in Tashkent, a separate video campaign is simpler: you see CPV and leads from YouTube, not “everything together”.
PMax as a second layer — when to launch Performance Max.
| Before | After | |
|---|---|---|
| One clip “everywhere” | An In-Stream test with a clear goal | |
| Only PMax “YouTube is in there” | A separate campaign for 2–3 weeks | |
| Bumper as the only format | Bumper after a working 15-sec clip |
03Creative: 6 seconds decide
Most people will skip. The first frames: what the service is, the city, what to do next. Not a logo for 4 seconds.
A working 15–30 sec frame for a service:
1. Seconds 0–3: the object / result (not stock office footage).
2. 3–8: the offer in one line + Tashkent / district.
3. 8–15: proof (a work shot, a review as on-screen text — no made-up numbers).
4. End: phone / “request on the site” / QR, if it fits.
Language as the client speaks. Subtitles: sound is often off.
Policies: medicine, “best in the city”, cure guarantees — risk of disapproval. “Book a consultation” is more honest than a miracle.
Shooting “like Reels” is sometimes cheaper than a studio. AI video is a separate conversation: AI video for Reels.
| Before | After | |
|---|---|---|
| 45 sec “about the company since 2010” | 15 sec offer + CTA | |
| No subtitles | Text on screen | |
| Stock New York | Your workshop / salon / yard |
04Where to send the click
Not to a homepage “about everything”, and not only to the YouTube channel header if the goal is a lead. Better: a service landing with a form and `tel:`.
On the channel: a header link, a pin, a playlist. But the ad should go to a page with the offer, or you cannot count CPL.
Conversions are the same as in Search: form, call, Telegram — conversion setup. Do not set Primary to “10-second view” if you pay for leads: Smart Bidding will start buying views.
| Before | After | |
|---|---|---|
| Click → homepage | Click → service page | |
| Primary = view | Primary = lead / call | |
| CTA only “subscribe” | CTA “estimate / book” |
05Test budget — a guide, not a price list
The figures below are a soft order of magnitude for SMB in Tashkent, not a lead guarantee and not a UZNEO price list.
To see if the creative is alive, you need enough views and clicks, not two days at $3. A practical test corridor: $8–20/day, 14–21 days, one clip, one geo (Tashkent or the service area), one goal.
If the daily money only buys a dozen views — there will be no conclusion. Two weeks of a normal test beat a month of “dripping”.
Compare not CPV with a neighbour, but leads and cost per lead against Search. YouTube often costs more per lead, but builds recognition. Then it is the top of the funnel, not a Search replacement.
| Before | After | |
|---|---|---|
| $3/day “to check YouTube” | 14 days with a sufficient daily budget | |
| 5 clips at once | One clip, then A/B | |
| KPI = views | KPI = lead / call |
06Audiences: not “the whole country 18–54”
For a service: service geo, language, interests carefully (Google does not read minds). Remarketing people who were on the site is a strong second layer when the pixel/tag is live. How lists work for Search — in remarketing; for video the same logic: do not push cold reach while there is no visitor base.
Exclude your staff and junk placements if you see a clear drain — without paranoia.
| Before | After | |
|---|---|---|
| Uzbekistan 18–54 | Tashkent + service area | |
| Only “business” interests | Site remarketing first, if there is traffic | |
| No placement exclusions on a drain | Exclusions from the placement report |
07How to read the test on days 7 and 21
Day 7: did it pass review? Is CPV not in space? Are there clicks to the site? If zero clicks with normal views — a weak CTA or the wrong offer.
Day 14–21: are there leads / calls from the campaign (UTM, call tracking). If only views — either keep it as a small brand layer, or pause and strengthen Search.
Do not compare CPV with “how a blogger does it”. You have a service and a lead.
| Before | After | |
|---|---|---|
| Verdict on day 2 | A 14–21 day window | |
| “Lots of views = success” | Leads and CPL | |
| Run forever with no review | Decision: scale / pause / remarketing only |
08Launch checklist
☐ 15–30 sec clip, offer in the first 3 sec, subtitles
☐ Landing with a form / phone, UTM on the ad URL
☐ Primary conversion = lead, not a view
☐ Geo = service area
☐ 14-day budget agreed
☐ Separate video campaign, not only PMax
☐ Policies: no forbidden promises
☐ Reply SLA if messages arrive
Help with traffic — Google Ads and online promotion.
Summary
YouTube for services in Tashkent is a separate test with a clip, a landing, and a 14–21 day budget — not a “checkbox inside PMax”. Urgent demand is more often closed by Search; video is warming and trust, if the offer is visible in 6 seconds. Start with one clip and one goal. To calmly work out whether you need this layer — UZNEO in Google Ads setup.
Launch ads without burning budget?
Google Ads with anti-fraud built in. Budget estimate and forecast in 24h on Telegram.
Related services
More articles
8
Advertising22 min
How to set up Google Ads in Uzbekistan
Advertising20 min
How Much Does Google Ads Cost in Uzbekistan in 2026
Advertising18 min
Yandex Direct or Google Ads: what to choose for business in Uzbekistan
Advertising22 min
How Much Ad Budget Do You Need by Niche in Uzbekistan
Advertising18 min
Google Analytics and Yandex Metrica: setting up analytics for business in Uzbekistan
Advertising18 min
UTM tags: what they are, how to set them up, and why analytics needs them
Advertising18 min
How to Protect Your Ad Budget from Click Fraud and Empty Clicks
Advertising17 min
Telegram