Skip to content
Mon–Fri 09:00–18:00
Conversion Optimisation — 17 min read

How to increase conversion rate: 21 proven tactics

Doubling traffic doubles the budget; doubling conversion happens on the same budget. Twenty-one tactics from measurement to cart abandonment, first screen to testing culture — each with what to do, why it works and how to measure it.

Burak Arda Özgül18 September 202617 min read

Doubling traffic doubles the ad budget; doubling conversion happens on the same budget. In the GYMWOLVES case sales rose 12× in three months, and the source was not a new channel: the data flow was repaired, the funnel rebuilt, the product pages improved; session duration tripled and engagement rose 8×. Same budget, different funnel.

Everyone typing how to increase e-commerce conversion rate into a search box is asking the same thing: how do we get more orders from the visitors we already have, how is the sales conversion rate lifted. This article is the implementation catalogue for that question. We have covered what CRO is and how the audit is run in separate articles; here there is only what to do.

Twenty-one tactics, grouped into five stages, each told in three lines: what to do, why it works, how to measure it. We did not go down to "change the button colour"; every tactic is a testable hypothesis. Wherever we use a number, its source sits next to it: our own cases or a named industry compilation.

Where do you start increasing conversion rate?

With measurement. However good the idea, a change made without knowing how many visitors drop at which step is a guess. That is why the first three tactics are about data, not the interface.

1. Build the funnel in GA4, event by event

What to do: define product view, add to cart, begin checkout and purchase as separate events in GA4 and build the funnel report from those four steps. Why it works: conversion rate is not one number but the product of four step-to-step rates; the weakest step is where the cheapest gain lives. How to measure: track each step-to-step rate weekly; no tactic starts before a baseline is taken.

2. Track the seven critical pages as separate pages

What to do: put the home page, the two busiest categories, three product pages and the checkout step on a separate watchlist, and read their exit and progression rates instead of the site-wide average. Why it works: orders pass through these seven pages, while the site average carries the noise of blog and about pages. How to measure: exit rate per page and progression rate to the next step; the GAP analysis article explains how the seven are scored.

3. Turn on session recordings and heatmaps on the funnel pages

What to do: collect session recordings and heatmaps only on the funnel pages, and watch the behaviour behind the drop-off number. Why it works: the number says where visitors drop, the recording shows why; together they produce hypotheses. How to measure: each week, watch a handful of recordings from the step with the highest drop and note repeating behaviour (back button, dead clicks, form abandonment).

Which 4 things on the first screen decide conversion?

On the first screen the visitor must understand what you sell, why to trust you and what to do next — and the page must load before they have to wait for any of it. Four tactics answer those four questions.

4. Reduce the value proposition to one sentence

What to do: the headline on the first screen says in one sentence which problem the product solves and for whom; slogans and campaign copy move down a line. Why it works: in the first seconds the visitor decides whether the page is for them, and leaves a page where they cannot decide. How to measure: exit rate on the first screen and time to first scroll; two headlines are compared in an A/B test.

5. Bring LCP under Google's threshold

What to do: bring the load time of the largest content element (LCP) under Google's Core Web Vitals threshold of 2.5 seconds, tackling image size, third-party scripts and server response time in that order. Why it works: speed is not an interface feature but a precondition of conversion; no tactic works on a page that has not loaded. How to measure: mobile LCP in PageSpeed Insights and the conversion gap between fast and slow sessions in GA4.

6. Move the trust signal to the first screen

What to do: the returns policy, secure payment, a real customer count or a recognisable client logo appears on the first screen just under the headline, not at the foot of the page. Why it works: the first question of a price-comparing visitor is "can I trust this site", and if the answer does not arrive before the first scroll the comparison continues in another tab. How to measure: progression rate from the first screen to a category or product; a with/without test of the trust block.

7. Leave one primary call to action

What to do: one dominant call to action remains on the first screen; "see products", "offers" and "contact us" do not sit side by side with equal weight. Why it works: equally weighted options raise the cost of deciding; a visitor shown one direction moves. How to measure: click-through rate on the primary call to action and the click distribution of the first screen on the heatmap.

What should change on product and category pages?

The purchase decision is made on the product page; the category page is the road to it. Product page conversion is the most fragile link of the funnel in most stores, and five tactics treat the two together.

8. Show the photograph in its context of use

What to do: next to the white-background shot, add frames showing the product in its setting, its scale and its detail; on mobile the first image shows the product in use, not alone. Why it works: the missing sense of a product that cannot be touched on a screen is closed by the image; uncertainty about scale and texture is a reason to return or give up. How to measure: add-to-cart rate from the product page and interaction rate with the image gallery.

9. Put social proof beside the product and user content in the gallery

What to do: review count and rating sit right next to the price, customer photos in the gallery, and named reviews in the body of the page rather than inside a tab. Why it works: another customer's proof is more convincing than the brand's promise; the UGC and social proof article explains why. How to measure: add-to-cart rate of sessions that saw the review block versus those that did not, broken down by review-count threshold.

10. Write stock and delivery information under the price

What to do: stock status, dispatch time and delivery cost are visible on the product page under the price, before the visitor goes to the cart. Why it works: when "when will it arrive" and "what does delivery cost" go unanswered, the visitor looks for the answer in the cart and abandons when the surprise appears there. How to measure: add-to-cart rate from the product page and exit rate on the cart page; a with/without test of the information block.

11. Show the price against an anchor

What to do: list price, discounted price and, where relevant, instalment or unit price (per item, per kilogram, per month) appear in one block; the price never stands alone. Why it works: price is perceived relatively, not absolutely; a price given no comparison point competes with a random anchor in the visitor's head. How to measure: add-to-cart rate and average order value read together across price-block variants; one should not rise while the other falls.

12. Make variant selection visible and error-free

What to do: size, colour and model are chosen with visible buttons instead of a dropdown; an out-of-stock variant is marked sold out rather than hidden; trying to add to cart without a selection shows guidance, not an error. Why it works: the variant step is the product page's most common snag; an invisible option and an unclear error turn into abandonment. How to measure: add-to-cart rate after a variant is chosen and the number of times the "choose a variant" error fires.

How do you reduce cart abandonment?

According to the studies Baymard Institute has compiled for years, the average e-commerce cart abandonment rate is around 70% — seven of every ten products added to a cart are not bought. Most of the reasons concern the checkout flow rather than the product, which is why the five tactics for reducing cart abandonment target that flow.

13. Make guest checkout the first option

What to do: continuing without an account becomes the first and most prominent option at checkout; account creation is offered with one click after the order is complete. Why it works: a mandatory sign-up puts a form in front of a visitor who arrived intending to pay, and every field of that form is a chance to give up. How to measure: progression rate from checkout start to completion of the first step; completion rates of the guest and member flows read separately.

14. Show the delivery cost before the cart, not in it

What to do: delivery cost and the free-delivery threshold are visible on the product page and in the cart summary, before the checkout step; the total does not change on the payment page. Why it works: an extra cost appearing at checkout is the most frequently cited reason for abandonment; the surprise damages trust more than the price does. How to measure: progression rates from cart to checkout start and from checkout start to completion, watched together when delivery information changes.

15. Offer payment options in the order the customer expects

What to do: options such as card, instalments, bank transfer and cash on delivery are listed on one screen in the order your audience uses most; the card form is a single step and open to autofill. Why it works: a customer who cannot see their usual payment route does not postpone the payment, they change the site. How to measure: progression rate from payment method selection to confirmation and the failed-payment rate per method.

16. Count the checkout steps and cut them

What to do: list every screen and every form field from cart to confirmation; the address is searched in one field, the billing address copies from the delivery address, phone and email are asked once. Why it works: every screen and every field is a drop-off point; completion rises as the number of fields falls. How to measure: abandonment rate per step and total time to complete checkout; both should fall when a field is removed.

17. Send the abandonment email within an hour, as one message

What to do: a recognised visitor who abandons the cart receives one email within an hour carrying the cart contents and a one-click return link; a discount is not in the first message but considered in a second to those who did not return. Why it works: intent stays warmest in the first hour after abandonment; a reminder brings back the customer who had already decided, while a discount teaches the hesitant one to wait for discounts. How to measure: return and purchase rate from the abandonment email; profit margin of the discounted and undiscounted variants read together.

How do you send traffic to the right page?

Sometimes what lowers conversion is not the page but the wrongness of the traffic arriving on it. Two tactics fix where ad and search traffic land.

18. Match the ad's promise word for word on the landing page

What to do: the product, price and promise in the ad appear in the same words on the first screen of the page it leads to; a campaign ad lands on the campaign's own page, not the home page. Why it works: every difference between ad and page is a "wrong place" signal to the visitor; at SOYLU AVM, once measurement was repaired and ad and page were aligned, $1.5M in revenue in the first six days was recorded. How to measure: landing page exit rate and conversion rate per campaign; the seven rules in the landing page article serve as the checklist.

19. Choose the page type by search intent

What to do: searches carrying words like "best", "comparison" and "price" go to a category or comparison page, searches carrying a model name straight to the product page, and searches starting with "how" to guide content. Why it works: showing one product to a visitor who wants to compare, or a category to a visitor who has already chosen, loses both. How to measure: landing page per query in Search Console and the organic conversion rate of the same pages in GA4, read together once the intent-page match is corrected.

Testing culture: which tactic comes first?

Applying nineteen tactics at once is impossible; deciding which comes first is itself a tactic. The last two are about ordering and validation.

20. Rank the tactics on an impact-effort grid

What to do: for every tactic, write the expected impact (visitors affected × the likely change in step rate) and the implementation effort (days) into one table; high-impact, low-effort items form the first sprint. Why it works: team energy goes to the most profitable job, not the most visible one; a small gain at the funnel's weakest step can bring more orders than a large gain at its busiest. How to measure: at the end of the sprint, write estimated impact next to measured impact; the estimation error corrects the next ranking.

21. Read the A/B test only when the sample is enough

What to do: calculate the visitors needed before the test starts, and do not stop it before it reaches that number and at least one full purchase cycle; two changes are never tested on the same page at the same time. Why it works: a test stopped early mistakes chance for a result; the A/B testing section of the CRO article explains why the sample matters so much. How to measure: the difference between variants on the target metric and whether the test reached its predetermined sample threshold; if not, the result is recorded as "unknown".

Who should do this: your team or an agency?

A question appears more and more often in search data: who are the best CRO specialists in Turkey for increasing conversion rates. Our answer is not a list of names but a distinction: a team that can build measurement infrastructure and a weekly testing discipline does most of these twenty-one tactics itself; for a team that cannot, the right question is not "which agency" but "what can the agency prove". The article on choosing a CRO agency gives the eight questions for asking for that proof.

At INDOLES, CRO consultancy begins with a diagnosis: the seven critical pages and the funnel are measured first, hypotheses go into the impact-effort table, then they are tested. The page of our conversion rate optimisation service sets out those steps and their deliverables; if the e-commerce infrastructure itself is the problem, e-commerce consultancy handles that layer. If you want to start on your own, the first step needs no budget: scan your store's seven critical pages with Diagnoo, and the report tells you which tactic comes first.

To sum up how we work as a CRO agency in one sentence: we do not buy new traffic, we find where the traffic you already have gives up, and we fix those points in order. The twenty-one tactics in this article are that order.

Frequently asked questions

What is a good e-commerce conversion rate?

No single figure is right; the rate varies by sector, traffic source, device and price level, and another store's number is no yardstick for yours. The useful comparison is your own baseline: today's step-to-step rates across the four funnel steps and where they move week by week. A store that watches its weakest step rather than an average always knows where to start.

How many weeks does conversion work take to show results?

Measurement and quick wins show in the first weeks: tactics such as moving delivery information forward, opening guest checkout and removing surprises read within a single purchase cycle. Changes that need testing speak once the sample fills; on a low-traffic site that takes weeks, on a high-traffic site days. The calendar is set by the site's visitor count, not the team's speed.

Why is mobile conversion lower than desktop?

Mobile visitors research more and complete less; filling forms, choosing variants and entering card details is harder on a small screen, and every difficulty is a drop-off point. The gap usually comes from the page rather than the user: a slow-loading image, a size hidden in a dropdown, an address not gathered in one field. Measuring and fixing the mobile funnel separately brings more orders than fixing the average.

Should you grow traffic or conversion?

The order starts with conversion. Conversion rate is the multiplier of all traffic: traffic bought before the funnel is fixed leaks through the same holes at the same rate, and ad cost enlarges the leak. Traffic added after the funnel is measured and the weakest steps are closed produces more orders for every unit spent. In the GYMWOLVES and SOYLU AVM cases the order was the same: measurement and funnel first, campaign second.

Is offering a discount in the cart abandonment email right?

Not in the first message. A reminder sent in the first hour after abandonment brings back the customer who had decided but got distracted, without a discount; opening with a discount leaves that customer unnecessary margin and teaches them to wait next time. A discount is considered in a second message to those who did not return, with the profit margin worked out. No decision is made before the purchase rate and margin of the two variants are read together.

Can a low-traffic site run A/B tests?

It can, but with few tests and large changes. The sample needed to measure a small difference takes months on a low-traffic site, so what gets tested is not button copy but the whole checkout flow or the layout of the product page. Where a test is not possible, before-and-after comparison, session recordings and funnel step rates give enough signal for a decision, provided the result is recorded as an observation rather than a certainty.

What tools does conversion work need?

Three layers are enough: an analytics tool such as GA4 for the funnel and step rates, a session recording and heatmap tool to see behaviour, and an A/B testing tool to compare variants. The brand of the tools does not decide the outcome; what decides it is correctly defined events and a weekly reading discipline. A wrongly defined purchase event blinds even the most expensive tool.

Which tactic delivers results fastest?

The tactics that remove surprises from checkout: moving delivery cost forward, opening guest checkout, deleting unnecessary form fields. None of the three needs a test, all read within a single purchase cycle, and because they fix the last step of the funnel their effect lands directly on orders. Starting the first sprint with these three gives the team both a quick win and a reliable baseline for the tests that follow.

What if conversion rises but average order value falls?

Read the two metrics together; if one rises while the other falls, the gain may be only apparent. A frequent cause is a lowered discount or free-delivery threshold: more people buy, each pays less. The deciding measure is not the number of orders but revenue per visitor; if that number is not rising, the tactic is rolled back or the threshold reset. Always write both metrics into the same table when price or delivery changes.

When does it make sense to hand CRO to an agency?

When nobody on the team will build the measurement infrastructure and read tests every week. Conversion work is not a one-off design job but an ongoing discipline; for a team that cannot sustain the discipline, outside support brings more orders than a team with tools. What to look at when choosing an agency is proof rather than promise: cases with numbers, the measurement method and honesty about test duration. The eight-question checklist sits in the article on choosing an agency.
B
AuthorBurak Arda Özgül

Founder · Brand Strategist & Creative Director

One of the rare people who keeps brand strategy and performance marketing at the same table. Builds the growth architecture of corporate brands; has worked alongside 40+ brands across Turkey, Europe and MENA.

Related service

Conversion rate optimisation (CRO)

Conversion rate optimisation is the work of getting more sales or enquiries from the visitors you already have; a measured lift is the only thing worth expecting from a CRO agency. INDOLES does not buy new traffic; it finds where the people already arriving give up, and fixes those points one by one.

View the service

Where do we start?

Three entry doors at three speeds. Pick the one that fits.

Submit brief