UBT for E-Commerce: How to Get Sales, Not Views

UBT in e-commerce does not start with a search for "free traffic". It starts with one question: how much does the store earn on a single delivered order? If $3.72 remains after the product, delivery and returns, while accounts, content and time consume $6.21, a million views will only increase the loss. That is why this guide starts with margin, then picks platforms, and only then produces content.

• UBT is not free — it is traffic without an ad budget.
• Physical-goods offers and brands are modelled differently, but the formula is the same: margin per delivered order.
• Minimum setup: 2 platforms, 3 content angles, 5 metrics and a 30-day test.

All amounts in this article are in US dollars. Values were converted at the October 26, 2026 exchange rate and rounded. The unit economics example is illustrative; plug in your own numbers.

What UBT for physical goods and e-commerce means

UBT stands for conditionally free traffic. You do not buy impressions and clicks directly; you earn people through organic recommendations, search and communities. "Conditionally" is the key word: content production, accounts, sometimes proxies and anti-detect browsers, people's time and product samples all cost money. UBT e-commerce is not "upload and forget"; it is media production plus operations.

In Russian arbitrage, "tovarka" usually means physical-goods offers (product arbitrage): dropshipping, cash on delivery, a short deal cycle and fast offer testing. E-commerce is broader: an own store, a brand, a marketplace, repeat purchases. The traffic mechanics are similar, but the economics are different.

No ad budget does not mean zero cost

Even if you do everything by hand, you still pay with your hourly cost, phone, editing, accounts and test samples. If you count only cash, UBT looks cheaper than paid traffic. If you count time and lost accounts, it is often more expensive — especially at the start, before you understand the platform and its moderation.

Why e-commerce math is different

In gambling and dating you often count a payout per registration or deposit. In e-commerce you count what remains after the product, delivery, payment processing, support and returns. Profit per order looks like this:

Profit = AOV × margin − delivery − payment fees − support − returns − UBT cost per order.

Contribution is revenue minus variable costs. It pays for traffic, fixed costs and profit, so compare cost per order with contribution, not with revenue.

Count delivered orders, not submitted ones. In cash-on-delivery offers, some orders are never confirmed or accepted; if you count them as sales, you can scale a loss.

Unit economics example

This is an illustrative model; plug in your own margin, delivery and returns. If contribution is $11.42 and the UBT cost per order is $24.83, you lose $13.41 on every order. For example, 100,000 views produced 1,000 clicks and 20 orders at $496.66 of costs — that is $24.83 per order. While this number is above contribution, scaling only increases the loss.

Low margin is the main killer of UBT. If margin is 20% on an $18.63 basket, only $3.72 remains before costs, and about $1.24–$1.86 after delivery and returns. Traffic at $4.97–$7.45 per order cannot save that math.

Product checklist for UBT

• Margin after product, delivery and returns is at least 35–40%.
• AOV and purchase frequency can pay for content and accounts.
• There is a visual or expert angle: before/after, demonstration, comparison.
• Returns and try-ons do not eat the margin.
• You know the LTV: one order or repeat purchases.
• Supplier, warehouse and support can handle order growth.
• There are no prohibited claims or regulatory restrictions.

Which platforms bring e-commerce traffic, not views

UBT is not one platform. Choose not by "where reach is bigger" but by "where the content fit and purchase intent are stronger". Here is a practical map:

In some regions, certain platforms are unavailable or restricted; check access in your geo and the platform's rules. If you send traffic to a marketplace, first check the marketplace and affiliate program rules: not every external link or promotion method is allowed. And do not spread yourself across five platforms: two platforms with a real content fit are almost always better than five where you repost the same clip.

For deeper platform breakdowns, see the guides to UBT on Pinterest and UBT on Threads.

How to choose two platforms

• You have visuals and search intent — start with Pinterest and YouTube (Shorts or long reviews).
• You have dynamics and impulse buying — TikTok and Reels.
• You have expertise and copy — Threads, Reddit, Telegram or your own blog.
• You have repeat sales — Telegram and email/SEO, not only short video.

Content that sells, not just collects views

The main test: a post or video should be useful even without a product link. If it only works as "buy here", the algorithm quickly finds a cheap audience that scrolls past.

• Problem → solution. "5 storage ideas for a small kitchen", with the product as one solution.
• Before/after. Cleaning, organisation, repair, garment care — visually provable.
• Unboxing + 3 use cases. Not just "look what arrived", but three specific ways to use it.
• Comparison of 3 models. For tech and expensive purchases: criteria, downsides, who each model fits.
• Mistakes and checklists. "5 mistakes when choosing a mattress" — useful and naturally leads to the product.
• UGC and reviews. Real only, with the author's permission. Fake reviews are a risk for both the platform and the brand.
• Seasonal collections. Pinterest boards and YouTube reviews for holidays, back-to-school, summer season.

One idea can become several formats: a 15-second clip, a carousel, a pin, a text post. But do not upload the same file everywhere: platforms may recognise recycled content. Make different covers, hooks and edits for each audience.

Content production can be accelerated with an AI creative factory, but a human still checks facts, rights and product quality.

Unit economics: five metrics before launch

Do not start with "how many views". Start with five numbers:

1. Cost per 1,000 views. Total UBT costs for the period / views × 1,000.
2. CTR to the store. Clicks / views. Shows whether the content hooks.
3. CR to order. Orders / clicks. Shows whether the promise matches the landing page.
4. Delivery and returns. Delivered / submitted, and returned / delivered.
5. Contribution per order. AOV × margin − delivery − payment fees − support − returns.

Then calculate cost per order: total UBT costs / delivered orders. If it is above contribution, you do not have a business, you have a hobby. If it is below, you can think about scale.

LTV is worth calculating too, but for physical-goods offers it often equals one order. For a brand with repeat purchases, LTV can justify a more expensive first order. Do not plug an "average market LTV" into your model — calculate it from your own base.

Tracking: without it, you are gambling

UBT is almost always multi-channel: several accounts, platforms and formats. If you cannot tell which account and which video produced an order, you are optimising an average.

• A unique link per account and content unit. Through a tracker or UTM tags.
• A promo code per creator/account. Useful when a link cannot be placed or gets lost.
• A postback from the store or affiliate network. So you see orders and deliveries, not just clicks.
• An attribution window. Decide in advance whether you count 1 day, 7 days or 30 days.

Minimum sheet: date, platform, account, content ID, views, clicks, orders, delivered, returns, revenue, contribution, costs, profit. If you run an affiliate offer, the tracker → offer → postback chain is covered in the tracker selection guide.

30-day launch plan

A minimum test is not one post; it is 10–15 content units per platform. One video can fail by chance; 10–15 show whether the angle and the platform work at all.

Red flags

• Views grow, clicks do not. The content does not bridge to the product, or the audience is wrong.
• Clicks exist, orders do not. The problem is the landing page, price, trust or promise.
• Orders exist, profit does not. Scaling will only increase the loss.
• Returns above 30%. Product, audience or expectations do not match.
• You need 100 accounts for one order. The model does not work; it is not "just warming up".
• Creative promises do not match the product. That is a ban risk and a customer-claims risk.

Risks and platform rules

UBT does not suspend the rules: spam, mass identical comments, fake accounts, engagement farming and misleading links lead to content removal and account bans. Read the official rules before launch:

• TikTok Community Guidelines.
• Meta Community Standards.
• Pinterest Community Guidelines.
• YouTube Community Guidelines.
• Reddit Content Policy.

There are consumer risks too: returns, warranty, delivery times, prohibited health and financial claims. If the product is regulated, check the requirements before producing the creative, not after a complaint. And do not send customer personal data to models or third-party services without consent.

One more risk is operational. UBT can bring more orders than you can confirm, ship and support. If the call centre and delivery are not ready, more traffic becomes more complaints and returns.

When UBT does not fit

• Margin below 30% and no LTV.
• A high return rate or a complex try-on process.
• An expensive considered purchase without demonstration and trust.
• A regulated niche with strict claim requirements.
• You need predictable volume "tomorrow" — paid traffic is better.
• There is no visual or expert angle, only "buy cheaper".

UBT does not fix a bad product, a weak landing page or an inadequate price. It can bring attention, but the offer closes the sale.

FAQ

What is UBT for an online store?

It is organic traffic from recommendations, search and communities where you do not pay directly for impressions. Costs go to content, accounts, tools and time.

How is UBT different from free traffic?

"Free" usually means no direct payment per click. UBT is conditionally free: accounts, content production and time cost money.

Can I start a physical-goods offer with no budget?

You can start without an ad budget, but not without costs. At minimum you need a phone, content, accounts and time; sometimes proxies, anti-detect tools and samples.

Which platform is best for physical goods?

It depends on the product. Pinterest suits visual and search-driven niches, Reels/TikTok suit impulse buys, YouTube suits reviews and tech, Threads/Reddit/Telegram suit expertise and community.

How many accounts do I need?

Start with the minimum needed to test content angles. A few accounts per platform are usually enough for a test; 100 accounts will not replace a bad offer.

How do I calculate UBT payback?

Calculate contribution per delivered order and compare it with the UBT cost per order. If cost is higher than contribution, the model loses money.

Does UBT work for brands?

Yes, especially Pinterest, YouTube, SEO and Telegram. But for a brand, long-form content and reputation matter more than mass identical accounts.

Which products do not fit?

Low margin, high returns, regulated categories, a complex considered purchase without demonstration, and products with no visual or expert angle.

Do I need a tracker?

Yes, if you want to know which account and content bring orders. At minimum, use unique links, UTMs and a sheet; better, a tracker with postbacks.

How long until the first sales?

It depends on the platform and product. Pinterest and YouTube may deliver a tail later; Reels/TikTok can be faster. A realistic test horizon is 30 days, not 3 days.

Bottom line

UBT for e-commerce is not "free traffic"; it is media production with unit economics. Start with margin and contribution, then choose two platforms with a content fit, then run a 30-day test. Sales, not views, should be the filter for every decision: if content does not lead to a delivered order with positive profit, do not scale it.

Begin with one product, one economic model and two platforms. If cost per order is below contribution after 30 days, you can grow. If not, change the offer, the angle or the platform — not the number of accounts.

Sources

Platform rules: TikTok, Meta, Pinterest, YouTube, Reddit. DERVALO.FORUM internal materials: UBT in TikTok, Reels and Shorts, TikTok warming and farming, UBT on Pinterest, UBT on Threads.