Introduction To Conversion Optimization

conversion optimization

conversion optimization sales funnel

Conversion optimization is a process whereby you change the content and organization of a website in order to increase the percentage of visitors that convert into customers.

This can be  a lucrative process as it helps you extract more value from your existing traffic.

If you’re wondering why more people don’t buy more of what you offer, or take the actions you expect them to, this article will give you a few ideas on how you can make changes to ensure they do.

Why Conversion Optimization Is Valuable

People are bombarded with messages.

The advertising industry is everywhere – it’s on your television, it’s on the internet, it’s in the mailbox, and it’s on your phone. Each message must compete for consumer attention. It must do so in an environment where many consumers have learned to block messages out. Supply and choice is bountiful, but demand remains limited by the size of the population, and by the average pay packet.

The consumer has a lot of choice.

In such an environment, we must make every visitor count. If we convert more visitors to paying customers, or encourage more to engage with us, we stand to be more successful than our competitors.

What Conversion Optimization Is Not

Conversion optimization is not just about e-commerce sites. It’s about all sites. When someone does something you want them to do, that’s a conversion.

Conversion optimization isn’t just about changing a landing page. It’s about looking at your entire approach, every step of the way. It’s about refining your offer to ensure it matches what people want.

Offer & Acceptance

We can make as many offers as we like, but we succeed when someone accepts our offer. Search engine optimization is mostly about the pre-offer and offer stages. We try and position a site against keyword search terms and achieve a high rank.

If we do our job well,  visitors will click on our link and arrive on our page. For many search engine optimizers, this is where their job ends.

Unfortunately, if that is all we do, we are unlikely to create as much value as is possible. A significant part of the value chain depends on what the visitor does next. The visitor could click-back within a few seconds. They may not feel the page is relevant to them. They simply might not like the look of it. They may have been distracted or got confused.

Conversion optimization is a process that aims to prevent these events from occurring quite so often.

On Average, 97% Of People Will Leave Your Site Without Converting

Search engine optimization brings in many potential buyers, but few actual buyers. Your mileage may vary, but a rule of thumb is that conversion rates across the internet are around 2-3%, meaning only two or three visitors per hundred will take our desired action.

Where do we lose the 97%?

Think of the web sales process as a funnel, with the widest end being the search engine results pages, and the narrow end being your shopping cart sales success page.

As the visitors move down through the funnel, they are getting closer to taking a desired action, but they are dwindling in number. They dwindle to the point where we’ve lost 97-98%, on average, by the time the 3% reach the narrow end – the checkout, or the desired action.

Why do we lose so many?

We lose people because we’re not meeting the customers needs. Conversion rate optimization helps determine what those needs are so that we can shape our offers and content to our potential customers.

In so doing, we increase the percentage of visitors who convert to customers, and thus make more money with the same amount of traffic.

Desired Action

Desired action is an activity we want the visitor to take.

For example, if we run an ecommerce site, we want the visitor to buy a product. If we run a news site, we might want the visitor to click on an advertisement. If we run a non-profit site, we might want someone to sign up for a newsletter. Most sites have multiple desired actions. When we undertake conversion rate optimization, we try to figure out exactly what the visitor wants, and ensure they get it.

We want to ensure their desired action aligns with ours.

The Increasing Cost Of Traffic

Optimizing our conversion rates can take the pressure off traffic acquisition spends.

Search traffic is becoming harder to get, due to more competition and less predictable search algorithms. It’s becoming more expensive to get people to the start of the funnel, so it makes sense to try and widen the funnel so we convert more people once they arrive.

Conversion Optimization is not as easy as simply tweaking a landing page. Like SEO, conversion optimization is an ongoing, iterative process that involves trying things out, measuring, and adjusting in order to produce a desired result.

The difficulty comes in knowing what the customer really wants. An offline salesperson can tell a lot about a person just by looking at them. They can refine and reshape the offer constantly by asking questions and countering objections.

This is harder to do online, but using conversion optimization, we can emulate this process.

Measure, Analyze, Change

Consider your existing conversion rate. Figure out what you want visitors to do, and examine how many visitors are currently doing it. This is called the baseline.

Once you establish a baseline, conduct an analysis. Look at your stats as visitors move through your site to see where you are losing them. Look at areas they are most interested in, and least interested in. Aim to get a snapshot of the current activity on a site.

Armed with this data, we move to the optimization phase. We change a little at a time. We might change the offer, the copy, the graphics, and the site structure. In practice, we’ll likely change all four aspects. We experiment.

Some changes will work. Some won’t. Some will be negative and send us backwards! The important thing to understand is that it is a process, and that process is iterative. Always be testing, as even subtle changes can produce significant results.

Google experimented with 41 shades of blue before they found the color that users responded to best.

Conversion Is A Process

One myth concerning conversion optimization is that all you have to do is test one page against another and repeat until you find the winning page design. Perfect! Now, watch the money roll in!

In reality, this seldom works. It doesn’t work because there are many factors that go into a conversion.

The conversion process is a series of steps. If any one of those steps falters, then we can lose customers, regardless of how well the design of a given page has been tested. In order to find out the problems and opportunities inherent in each step, we must use a systematic, repeatable process.

Conversion doesn’t just happen at the end. People don’t just click “buy now” and hand over their credit card details. As we’re seen, only 3% of our visitors, on average, will get that far. What about the other 97%? We’re losing the other 97% at various steps in the process. Each step is an opportunity to optimize conversion.

Goals & Steps

It’s useful to map out every possible goal and step if only to test whether you should eliminate some!

For example, a sales cart process could look like this:

  • Visitor clicks on order button
  • Visitor is asked to create an account
  • Visitor is asked for name
  • Visitor is asked for email address
  • Visitor is asked for physical address
  • Visitor is asked for phone number
  • Visitor is asked for mobile number
  • Visitor is asked if they want to go on mailing list

I’m sure you’ve spotted one problem straight away. Why does the buyer have to create an account before they can make an order? Our hunch might be that buyers don’t like creating an account at this point.

In order to prove our hunch, we should watch abandonment rates at this point in the process. If we see high abandonment rates, we might create an option that allows the user to place an order without setting up an account. We then see if this change results in higher conversions.

The buyer might like a painless ordering system that lets them buy, then asks them where they want the item delivered. It’s a subtle distinction. We glean the same information, and can use this information to set-up an account, but we haven’t inhibited the purchase.

We should then think about what fields are really necessary. We may require all those fields, but we should also ask ourselves if they are strictly necessary. If so, are they strictly necessary at this exact point, or can they be derived later?

The Value Of Different Options

Each option a user takes will have a different business value.

If our aim is to collect email addresses in order to build a mailing list, then we might give the email field a high value relative to other options, such as a search field. We might repeat the request for the email address on various pages, or feature the form prominently, whilst making other options less frequent or prominent.

Some optimizers feel that we should only give the visitor one goal. For example, the goal of landing page X is to result in a sale. Whilst this makes it clear for the business, it may not work well for the visitor. If the visitor wants to research information before making a decision, we lose them if we don’t provide the visitor a path to find this information.

It’s similar to the way businesses use KPIs to help track their performance. KPI stands for Key Performance Indicator. KPIs help you define and measure your progress towards your goals.

You may have a goal to “make more money”. Let’s refine that a little and make it more explicit. “The goal is to boost revenue by $500K this year”. We then work backwards from this goal and work out the steps needed to achieve it. Each step must be measurable and quantifiable.

Each option leading to a goal can be stated in micro terms i.e. “we want 10% of people to click on the shopping cart button”, as we’ve determined that the more people who click on the shopping cart, the more likely they are to progress to the next step, which is to fill out their credit card details.

Goals will, naturally, differ depending on the type of site. In e-commerce,indicators of success or failure may include the number of conversions, order values, and checkout process abandonment rate. They could also include the number of purchases people complete off-line i.e come into your store as a result of researching on your website.

Subscriptions and advertising-driven sites will have different types of goals and values. They will likely measure the number of visitors, the number of page views, the number of subscribers, the number of subscribers who cancel and the average length of subscription. These values could lead to the same overall goal “to boost revenue by $500K this year”

Once defined, we know what we have to change to achieve our goals. We also have a means to measure the effect of the changes.


If we know the value of an order, we have a clearer idea of how much to pay for our advertising. Is $0.50 cents too much for a click? $2.00? $8.00? How much should we spend trying to optimize for position #1? How much is position #1 worth?

Let’s say a unit we sell has a margin of $100. If we get 1,000 visitors, and convert at 3%, that means we sell 30 units, which means we make $3000. If we spend $3000 in advertising, we break even. We could spend up to 0.33 cents for those clicks. Anything below this figure, we make money.

This is a very simplistic way of looking at it, of course, as we have other costs, such as the total campaign cost including design and optimization, and the lifetime value of the customer, but it’s a way to illustrate how we can set and manipulate budgets once we know the conversion rates. It’s also a way to show how profitable conversion rate optimization can be if we increase the rate.

If we doubled the conversion rate, we’d see significant improvement in the bottom line. Using the same example, a unit has a margin of $100. If we get 1,000 visitors, and convert at 6%, that means we sell 60 units, which means we make $6000. If we spend $6000 in advertising, we break even. We could spend up to 0.66 cents for those clicks. Anything below this figure, we make money.

Bounce & Exit Rates

The “bouncers” are the people who land on your page, but click back. Exit rates are determined by the number of people who get someway through the conversion process, yet abandon it at some point. These are the 97%ers. These are the people we want to try and convince to stay.

Bounce rates have a lot to do with relevance. If people click back immediately, it’s because they don’t feel the page is relevant to them. They may have clicked on a PPC ad that led them to believe one thing, but when they arrived, they were shown something else. Or, perhaps, those two messages were in sync, but the page itself didn’t appeal to them, or wasn’t clear.

We can do a lot experimentation around bounce rates. We could change the PPC text. We could change the design of the landing page. We could change the copy of the landing page.

Exit rates give us a good idea where to start optimizing. If people get to step X of the shopping cart process, but abandon the process in high numbers at that point, we can safely assume there is something wrong at that point. We could change this to make it easier, more transparent, we could address security concerns, or perhaps place a chat application at this point, prompting people if they appear to need help.

The Customer Is King

The concept at the heart of conversion optimization is that the customer is King. The design, the copy, the process must all be aligned with your customers. Failure to do so will result in click-backs and loss of revenue.

But how do we know what the customer wants?

We could guess. Web designers tend to design what they think looks good, which is fine, but measurement drives conversion rate optimization. We’re not interested in what designers *think* might work.

We need to prove what actually works.

Who Is Using Our Site? What Do They Want?

One means used by marketers to get a better idea of what their users want is by using “personas”.

Personas are archetypes of customers. It’s a way to put yourself in someone elses shoes and imagine seeing the world through their eyes. This is important, because our customers aren’t all the same. They will likely be buying for very different reasons.

Imagine you run a car yard. A 20-something female walks onto the lot. She wants to buy a reasonably priced late model car. Something not too expensive, but something that looks great. She’s not overly interested in specifications or performance, but it can’t be a slug. Fast is good!

Minutes later, a middle aged man walks onto the lot looking for a vehicle he can use for his construction business. He’s pretty sure he wants a pickup. Or maybe an SUV. He doesn’t care about price. He also wants something he can take offroad in the weekend.

Two very different customers who both want something you have. You need to talk to these customers differently. This is easy to do offline, where you can size them up, watch their expressions, ask questions, listen to their responses, and interact. This is harder online where you tend to track what they do by observing data.

If we create personas for these different people, it can make it easier to determine how the site should look, and where we should place certain features and messages.

To create a persona we detail a personas specific characteristics. The young woman becomes Zoe, a 25 year old advertising executive. She is interested in fashion and travel. She doesn’t have much disposable income, but has got a good line of credit. She is concerned about price, but she is most concerned about appearance.

Similarly,  Warren is a 55 year old builder, married, with two kids. He’s well off. He’s not spendthrift. He likes things to be solidly built and dependable.

What Personas Should We Choose?

We need to try and understand our customers. We can do this using surveys. We can monitor feedback. We can study the language used in the emails they send us. We can buy market research data, such as demographic information. We can tap into our experience of previous customers.

The field of psychology helps us build credible personas. You may be familiar with Jungian archetypes and Myers-Briggs Type Indicators. These provide an intellectual basis for categorizing people into groups of people who will behave differently in the same situations, based on the way they experience and respond to the world.

In terms of conversion optimization, we should be aware that there are impulsive types, cautious types, and logical types. They will likely respond to different triggers.

For example, we don’t want a shopping cart to get in the way of an impulsive person (see Amazon’s solution in the form of 1-Click ordering and in-app purchases by games manufacturers). On the other hand, we need to provide enough reassurance for the cautious person. Examples include security information, a returns policy, and brand awareness. Logical types will require sufficient information in order for them to make a considered judgement.

There are many cross-overs, of course. A person may be logical in one context, yet impulsive in another. The important thing is to test your website with each personality type in mind and see if it creates problems for each type. As we build up a knowledge base of our customers, we can constantly refine our offer and positioning.

Let’s revisit our car yard example. Why is this car buyer coming onto a lot? Why did they choose this lot, and not another? What do they expect to see? A lot of cars? Few cars? Cars displayed inside? Outside? More of one type of car than another? What are the questions this persona is likely to ask? What are their likely objections? What would make this persona go away? I’m sure you can think of many questions once you have a specific persona in mind. You can then switch persona, and look for areas where the questions, and responses, will be similar, and where they change.

In terms of a web site, each persona may raise questions about shipping. Is free shipping important to this persona? To all personas? If you determine it is, then you might want to make a prominent feature of free shipping.

Another issue that may come up is trust. You might find it’s not as important as the persona browses and gathers information, but becomes a big issue at checkout. Your trust messages could be hinted at through your site, but highlighted at checkout time.

Design a range of personas, then step through your website. Look for areas that seem to fit well with one persona, but grate with another. These areas need work.

We can change the layout of these areas. Change the graphics. Change the copy. Experiment to see if we can find areas that don’t grate against different personas. We can test in a live environment to see what visitors actually do. If we notice an uptick in one customer type, or a downturn in another, we can tweak and adjust accordingly. We need to think about ways to gather demographic information to better inform our process.

Personas aren’t a magic bullet. Will Warren and Zoe really shop differently online? Or will they both respond equally well to an easy-to-use website? Are these personas really rooted in reality i.e. hard data, or are they imagined?

I think personas are a useful means of approaching site design and site structure in that they force us to look at things differently. They force us to test assumptions.

Testing assumptions lays at the heart of conversion optimization.

Value Proposition

No matter how good your usability, your personas, or how easy your shopping cart is to use, your site won’t succeed unless you have a clearly articulated value proposition.

A value proposition is a promise of value to be delivered. The customer must believe they will derive this value. A value proposition can be weak or strong. A weak value proposition might be when the customer can easily find the same value, elsewhere. A strong value proposition might be when the customer can’t get what you offer anywhere else. You may offer lower prices. You may offer something that improves their lives markedly. Whatever your value proposition is, ensure it is clearly articulated.

Keep in mind that your value proposition will be relative. Customers will compare your value proposition with other sites. So, part of your task in optimization is to compare your offering against those of your competitors. Look for opportunities where your competitors may be catering for one type of customer, but ignoring another i.e. they’re talking to the Zoes, but ignoring the Warrens. This presents a market opportunity.

Value propositions are linked to the motivations of your customers. Some customers will be motivated to solve a problem i.e. they have something in mind they wish to buy. Other customers will be motivated by the need to gain information. The value proposition you offer may be different for each type of visitor.

Your site needs to articulate your value proposition to as many different types of customer as possible.

Establishing Credibility

The visitor must deem your site to be relevant.

SEOs know the value of relevance i.e. the keyword term must match the landing page. In conversion optimization, we extend this idea all the way through the site. There should be a consistency from search result to shopping cart. If you advertise the cheapest prices, then you must display them, and the final shopping cart price must demonstrate it. Fail to do so, and you’ll lose the visitor who clicked through on a promise of low prices.

A visitor may look for signs of social proof. The danger in buying is that it presents a risk. There is the risk the retailer won’t deliver as promised. There is social risk that the buyer will feel let down by their purchase. Their friends might not think this is a good idea. They seek validation. This is why providing social proof is a powerful tool that aids conversion. Who else has purchased this? Are they my peers? Strength in numbers may be a somewhat irrational validation, but it’s still a powerful conversion technique.

Other methods of establishing credibility include membership of recognized trade organizations, reputation in the market, and increasing brand awareness.


Entire books could – and have been – written about design. In terms of conversion optimization, there is only one way to know if your design helps, or hinders, your visitors.

Test. Make a change. Retest.

I suspect you already know that making the design easier to use will pay dividends. The tricky part is working out exactly where a design can be made easier, especially if the site is already easy to use and free of obvious problems.

This is why it’s important to watch how other people use your site, especially those who aren’t already familiar with it. There are various services available that will present your design, and ask users what they think your site is about. There are various services that track user eyeball paths. And your analytics, of course.

If you spot areas that cause problems, make incremental changes, then retest. It’s best to make minor changes, as then you can isolate the problem without risking introducing others.


You may have heard of split/run testing, also known as A/B testing. Split/run testing is when you test different designs, side by side, and see which design works best.

For example, you may create a landing page where you want someone to sign up for an email list. Do we put the request for sign up at the bottom, after we’ve explained everything? Or should it be the first thing people see? Should it be red? How about blue?

We don’t know.

So we test the different layouts against each other and see which one works the best. We can change every variable, including copy, headings, images, placement, and the number of times we restate the offer. Rather than rely on our hunches, we test.

Similar to A/B testing is multivariate testing. This testing allows you to  test more than one component at once. Some websites benefit from constant and continuous optimization as visitor response to creatives and layouts differ by time of day/week ,or even season.

A/B is obviously the simpler of the two. It’s probably a good idea to start with A/B testing, and then introduce multivariate testing when you’re trying to squeeze every last bit of incremental performance out of your pages.

More Than Tweaking Pages

As you can see, conversion optimization is not just about tweaking landing pages or finding the right copy. It starts with a detailed business analysis of strengths and weaknesses and flows all the way through your approach and site.

Don’t settle for 3%!  🙂

VN:F [1.9.22_1171]
Rating: 7.0/10 (19 votes cast)
VN:F [1.9.22_1171]
Rating: +2 (from 6 votes)
Introduction To Conversion Optimization, 7.0 out of 10 based on 19 ratings

Leave a Reply

Your email address will not be published.

Want to try SEM Rush? click here for a free 14 day trial.