Your conversion rate is the share of visitors who do the thing you want. That might be buying, booking, calling or sending an enquiry. So the first job is to pick that thing.
If your goal is for people to read the text on your site, you need a new goal.
Peep Laja, founder of CXL, in 13 Ways to Increase Your Conversion Rate Right Now
Work out your number
Choose one main action for your website:
- An online shop: a purchase.
- A plumber, dentist or salon: a call, a booking or a sent form.
- A coach or consultant: a booked call.
Then do the sum. Divide your actions by your visitors and multiply by 100. If 1,000 people visit in a month and 20 get in touch, your rate is 2%.
Here's where to find those numbers:
- On Shopify: go to Analytics, then Reports, and open Conversion rate breakdown. It shows how many visits reach each step, from the cart to the finished order.
- In Google Analytics: tell it what your main action is. Go to Admin, then Events under Data display, and click the star next to that action. Google calls this a key event. Purchases are already marked for you.
Count phone calls too. If most customers ring you, your reports will miss them and your number will look worse than it is. For one month, ask each new caller how they found you. A sticky note by the phone will do.
Now work out what a better rate is worth. Say you get 1,000 visitors a month and 20 enquiries. If your rate goes from 2% to 3%, you get 30. That's 10 more enquiries from the same visitors. If you win half and a job pays $400, that's $2,000 more a month. Put your own numbers in. The answer tells you how much time and money this deserves.
Watch your profit while you do this. A huge discount will lift your rate and can still lose you money.
What is a good conversion rate?
It depends on what you're measuring. A whole shop and a single page aren't the same thing, because many visitors to a shop are only looking.
| What you're measuring | What's typical | What's good |
|---|---|---|
| A whole online store | About 1.3 sales per 100 visits (1.3%) | Above 2.7 per 100 beats 4 out of 5 stores |
| A landing page, which is one page built for one ad or offer | About 6 or 7 actions per 100 visitors | 10 or more per 100 |
| A whole site for a local service | No number you can trust | Better than your own last month |
The store numbers come from Littledata's count of 438 Shopify stores. The landing page numbers come from Unbounce's look at 41,000 pages. Those actions include sign-ups as well as sales, and a sign-up is far easier to get.
Your best comparison is your own number from last month. It has the same customers, the same prices and the same website.
Find where people stop
Fixing the wrong thing is how owners lose a month. Use this table to choose.
| What you see | What it usually means | Start with |
|---|---|---|
| Fewer than a few hundred visitors a month | You need more of the right people first. Your rate won't mean much yet. | Fix 1, then get more visitors |
| People leave within seconds | They're the wrong people, or they can't tell what you do | Fixes 2 and 4 |
| People read but don't click | They don't trust you yet, or there's no clear next step | Fixes 3, 5 and 6 |
| People click, then give up on the form or checkout | It's too long, or the price was a surprise | Fixes 6 and 7 |
| Computers do fine, phones do badly | Something is awkward on a small screen | Fix 8 |
| You get enquiries, but few become customers | You're replying too slowly, or attracting the wrong people | Fixes 1 and 4 |
A website not converting anything at all often has something broken, so check fix 1 today.
If you can't tell which row is yours, these three things will show you. All of them are free.
Watch real visits. Microsoft Clarity is free and records what visitors do on your site. Install it, wait a week and watch 20 visits from phones. Look for people tapping things that aren't buttons, scrolling up and down in search of something, or quitting halfway through a form. Mention the recordings in your privacy notice.
Watch five people. Find five people who could be customers but don't know your business. Hand them a phone and say, "You need what we sell. Show me what you'd do." Then keep quiet, which is the hard part. Researcher Jakob Nielsen found that after about five people, you mostly see the same problems again.
Ask your customers. Ask each new customer one question: "What almost stopped you from getting in touch?" Write down their exact words. You'll learn what's getting in the way, and you'll also get better wording for your page.
The fixes, in the order to do them
They start with what's cheapest and most likely to matter. Skip any that your table row says you don't need.
1. Make sure it works, and that someone answers
Send an enquiry through your own site from your phone. If you run a shop, buy your cheapest product. Then call the number on your site.
Check that the enquiry arrived, and look in your spam folder as well. Contact forms have a way of sending everything to an address someone set up years ago. Business may not have been slow. Your old web designer may just have been very popular.
Then reply quickly. A person who fills in your form may be filling in two others. Aim to reply within minutes during opening hours. After hours, send an automatic reply that says when they'll hear from you.
2. Say what you do on the first screen
Show your homepage to a friend on a phone for five seconds, then take it away. Can they tell you what you do, who it's for and what to do next? If not, change the headline. Only promise what's true.
| Instead of | Try something like |
|---|---|
| Welcome to Smith & Sons | Emergency plumber in Dayton. At your door within 2 hours. |
| Your trusted partner in wellness | Physical therapy for back and knee pain. Appointments this week. |
| Transforming lives through coaching | Career coaching for managers who want a new job this year. |
| Premium handcrafted goods | Slim leather wallets that fit a front pocket. |
Next, ask someone outside your trade to circle every word they wouldn't say out loud. Replace those words. You see your site every day, so you're the worst person to judge it.
3. Give each page one main next step
Decide the one thing you want people to do on each page. Make that button the easiest thing to see, near the top, and repeat it lower down on a long page.
Use words that say what happens next. "Book a free 15-minute call" tells people far more than "Submit". Then take away things that compete with it, such as sliding banners and long menus. On a page made for one ad, try removing the menu altogether.
4. Send people to the page they came for
If your ad says "furnace repair", send people to your furnace repair page and not your homepage. Otherwise they have to hunt for it, and many won't.
Then look at where your visitors come from. In Google Analytics, go to Reports, then Acquisition, then Traffic acquisition. If one source brings plenty of visits and no actions, stop paying for it. Your rate can rise just from turning off the wrong visitors.
5. Put proof next to the button
People want to know that someone like them took this step and was glad. So put two or three real reviews beside your button or form. Use the person's name, their town and what they bought. Add real photos of you, your team and your work.
You need fewer reviews than you think. Northwestern University researchers found that shoppers were far more likely to buy a product with five reviews than one with none. After about five, each extra review helped less. They also found that people were most likely to buy at 4.0 to 4.7 stars, and less likely as ratings got close to a perfect 5. So one bad review isn't a disaster.
Ask for a review on the day you finish the job, while people are still pleased. Never buy reviews or write your own. Fake ones are banned in the US.
6. Answer the worry before they ask
Every visitor has a question they won't phone you to ask. So answer it on the page.
| The worry | What to add |
|---|---|
| How much will it cost? | A price, or a range such as "most jobs cost $200 to $400" and what changes it |
| How long will it take? | A real delivery date or the next open appointment |
| Will it work for me? | Who it's for, who it isn't for and one example like theirs |
| What if it goes wrong? | A guarantee you can keep and a clear refund policy |
| Are you real? | Your address, phone number and a photo of an actual human |
Start with price. When Baymard Institute asked US shoppers why they gave up on a cart, extra costs like shipping and fees were the top reason among those who'd come to buy. Hiding the price rarely stops people caring about it. They usually just leave to find it somewhere else.
7. Shorten the form or checkout
Look at every box and ask whether you need the answer today. An enquiry form can often be a name, a phone number or email, and one line about the job. You can ask everything else when you talk.
Shops have the same problem. Baymard found the average US checkout shows about 15 boxes to fill in, when 7 or 8 can be enough. Let people check out as a guest, and offer Apple Pay or PayPal so they don't have to find a card.
8. Try it all on a phone
Do the whole thing on your own phone, with one thumb. Can you tap the phone number to call? Are the buttons easy to hit? Does the form bring up a number keypad for the phone number?
Don't assume phones simply sell less. In those same Shopify stores, the typical phone rate was a little better than the computer rate. So a poor mobile conversion rate usually means something on your site needs fixing.
9. Make it load faster
Go to pagespeed.web.dev and type in your address. Then look at the phone result.
Portent studied 20 websites and compared how fast their pages loaded. On sites collecting enquiries from other businesses, a page that loaded in 1 second got people to act about 3 times as often as one that took 5 seconds. Online shops saw a gap nearly that big. So if your pages take more than a few seconds on a phone, move this fix up your list.
Shrink your big photos first, then remove any apps and plugins you don't use. Speed is the part of website optimization that's easiest to measure, but a fast page that confuses people still won't sell.
When visitors need an answer right there
Everything so far improves a page that says the same thing to everybody. But some visitors have a question only they would ask. A patient wants to know if you treat her kind of knee pain. A shopper can't choose between three sizes. If nobody answers, they go.
That's why we built Possible, so take this with the obvious pinch of salt. Possible is an AI built on real sales psychology. It asks each visitor the right question at the right moment, remembers what they say and guides them to a sale, sign-up or booking. It's free to start. Here's what it looks like as an AI sales assistant for clinics and local services.
Fix a confusing first screen before you add it. The same goes for anything sold as ai conversion rate optimization, because no tool can rescue an offer people don't understand.
What to leave alone for now
- A/B tests. This means showing two versions of a page to see which wins. Say you want to spot a rise from 2 to 2.4 sales per 100 visitors. This calculator says you need about 20,000 visitors for each version, so 40,000 in all. At 2,000 visitors a month, that's 20 months of waiting. Make the change, compare the next month with the last and keep asking customers.
- A redesign. A new look on a confusing message is still a confusing message.
- Button colors. People who can't tell what you do won't be won over by orange.
- Fake countdown timers. Real deadlines are fine. A timer that resets when you refresh the page tells people what else you might fake.
- More ads. If your page isn't working, more ads get you the same result at a higher price.
Your first week
- Today: Test your form and your phone number. Work out your rate.
- Tomorrow: Find your row in the table. Install Clarity if you can't tell which one it is.
- Day 3: Rewrite your first screen and your main button.
- Day 4: Add reviews near the button and put your price or a range on the page.
- Day 5: Cut your form down. Try the whole thing on your phone.
- Days 6 and 7: Watch five people use your site. Fix whatever made them stop.
Then leave it alone for a month and compare your numbers. Check that the month you're comparing with wasn't your busy season.
Questions about increasing your conversion rate
How do I increase my conversion rate?
Find the step where most people stop, and fix that step first. For most small sites, that means a clearer headline, one main button, real reviews near that button and a shorter form. Change one thing at a time when you can, so you know what helped.
How can I optimize the conversion rate of my website?
Pick one main action, measure how many visitors take it, and watch a few real people use your site. Then work through the fixes in order. To optimize just means to keep removing whatever stops people.
Is a 7% conversion rate good?
For a whole online store, 7% is excellent. Only about 1 Shopify store in 10 gets above 3.6%. For a single landing page, it's about typical. For your business, it's good if it beats last month and you're making money.
Is a 12% conversion rate good?
On a landing page, yes. On a whole online store, it's so high that you should check your numbers first. Your report may be counting something other than real orders. A service business might reach it when most visitors already know the company, so look at where they came from.
Should I get more visitors or fix my conversion rate first?
Fix what's broken first, because that takes a day and costs nothing. After that, if you have fewer than a few hundred visitors a month, get more of the right people. You can't learn much from a rate based on 40 visits.
How long does it take to raise a conversion rate?
Most of the fixes here take an hour or two each. Seeing the result takes longer, because you need enough visitors for the numbers to mean something. On a small site, judge a change after a full month.
Can a higher conversion rate lose me money?
Yes. Big discounts can bring more buyers and less profit. A form that's too easy can also bring in more people who were never going to buy. Watch your sales and profit alongside your rate.