Paid advertising can be one of the best ways to scale a business. As a long time veteran organic marketer, specializing in SEO, it took me a long time personally understand and value this.

Now I believe that with budget available to invest in it to learn how well it works, it can be a great way to scalably grow your business and do it profitably so you can then invest back into other, dare I say slower, channels like SEO and content marketing.

When measuring paid advertising performance, there are two main terms to know:

  • ROI – Return On Investment which takes into account all expenses that go into advertising including not only ad budget but also salaries of ad buyers or agency costs, costs associated with design and copywriting, and more. For example, if all costs associated are $5,000 and $25,000 in revenue results, ROI is 5. A common ROI is 2-4x.
  • ROAS – Return On Advertising Spend which purely looks at ad spend and resulting revenue. If ad spend is $2,500 and resulting revenue is $25,000, ROAS is 10. A common ROAS is 3-5x.

A full writeup on the difference between ROI and ROAS is here.

There are many more ways to improve ROI than there are ways to improve ROAS. This article will focus on improving ROAS itself.

12 ways to improve ROAS

The tactics and strategies you use to improve your ROAS will vary based on your company’s buyer and offer (ecommerce vs SaaS vs professional services, for example). 

Below is a list of the most commonly available strategies to improve ROAS, with the goal that at least a few are relevant to your business and you can take action to see a change in your fortunes (literally!).

Optimize landing pages

When you’re running ads, you’re literally buying traffic from one site and getting it to yours. The more traffic you buy and then convert, the more money you’re going to make.

This is why the first place to begin when trying to improve ROAS is your landing pages. 

  • Do they load fast?
  • Is the offer clear?
  • Does what your ad talks about translate to your landing page and what the prospect was expecting to see or be able to do?

The most common highest leverage landing page optimizations to make include a headline that speaks directly to their need, copy that takes them along a journey to being ready to buy, and an easy way to get started.

Without a clear headline that makes them stay, copy that helps them see themselves and their need, and a clear call to action your conversion rates and thus your ROAS will suffer.

Improve page load times

Did you know that website conversion drop by 4.42% on average for each additional second of load time between seconds 0-5? Source

At a 1 second load time, conversion rates are on average 31.79%. At 5 seconds, they’re 9.68%.

This right here is straight math. If you improve your paid load time from 5 seconds to under 1 second, you should see ~3x the conversions. If you improve from 5 to under 2, you’ll see 2x. If you improve from 5 to under 3 you’ll see a 50% improvement.

NBED IMAGE https://www.portent.com/blog/analytics/research-site-speed-hurting-everyones-revenue.htm

Even a 50% improvement in conversions would be massive for your business, right?

Some tactics you can use to improve your page load times are:

  • Properly size images (use an image plugin in WordPress to do this easily)
  • Cache and offload resources wherever possible (using a plugin like WProcket)
  • Defer unnecessary scripts to the bottom of the page, so they don’t block initial loading

Improve page readability

Readibility is of utmost importance these days on the web, with screen sizes and resolutions varying while the population also gets older. If you’re a B2B company especially, you need to pay attention to the devices on which your audience are viewing your pages, because different screen sizes require different approaches.

That said, the current recommended font sizes on desktop should be between 16-18px (1.6-1.8em). Headings should be ~2x the size of body text, and subheadings scaled down from there to keep contrast.

On mobile, font sizes should be at least 16px with headings ~1.3x the size of body copy. Also make sure to allow line spacing so that it’s natural to read what’s on the screen.

Improve page copy

Page copy. Let’s talk about page copy.

So many companies write copy that is terrible. That’s just a fact.

Their copy is self-serving, bland, and generally just doesn’t convert because it doesn’t actually speak to the ideal potential customer.

Even if the copy does a decent job of speaking to the audience, it’s then full of typos and unclear language that hurts conversion rates substantially.

Don’t drive traffic to a page with typos, especially since we know that up to 85% of millennials are less likely to buy a product advertised with errors (source).

Then make sure your copy is talking to your customer about their pains and wants instead of your offer. Go through and rework any instances about “our” offer/platform/whatever it is that you’re selling.

Sell them a transformation, not an offer.

Change up your form

The common wisdom when testing a form to increase conversions is to remove form fields. Many studies show us that removing form fields can translate into an increase in volume of leads. This example from Marketing Experiments showed that in one study, reducing the form from 9 to 5 fields increased conversions by about 30%, from 10% to 13.4%, and dropped their Cost Per Lead (CPL) by about $10.

NBED marketo-form-field

That said, there are also studies such as this one from CXL that have shown opposite results. In that study, Michael Aagard conducted a test where they removed a third of fields and saw something interesting.

Well, it’s interesting to a conversion optimizer but not so much to a client. They saw 14% drop in conversion. 

Michael says:

“I removed all the fields that people actually want to interact with and only left the crappy ones they don’t want to interact with. Kinda stupid.”

They added back in the fields, changed up field copy, and then saw a 19.2% increase in conversions.

This is why you should learn best practices and see if they work for you, which means testing them in your own environment.

Clarify your offer

Sometimes people don’t buy because they’re not clear on what you are offering.

This is less common in ecommerce (though adding things like free shipping, or just clarity around shipping costs or taxes during the checkout process, can help) but extremely common for services businesses or even SaaS companies.

If you’re seeing lower conversion rates than you should be (for most sites, 2-3% is a good conversion rate), try the following:

  • Sell the destination, not the journey. As someone once said to me, “Sell Hawaii not the plane trip.”
  • Clarify what they get working with you. Add social proof, your depth of knowledge, testimonials, and that sort of thing.
  • Play with language around what happens next after they contact you.
  • Make it easy to contact you and that the initial consultation is free (if that applies).

One word of caution – I have found that it can be possible to explain TOO MUCH ahead of time. For example, on Credo’s main form I tried various ways of explaining what happens after one submits the form.

Every time I implemented something about them scheduling a call with me/the team, conversion rates dropped by about 20%. I’d remove it and conversions would come back up.

Make sure you’re measuring conversion rates to know if something like this happens. My intuition told me conversions SHOULD go up because visitors had more clarity, when in actuality it hurt conversions.

Always test, always measure.

Lower the commitment

Sometimes we ask too much of people before we build trust of them. As the old saying goes, “don’t propose marriage on the first date.” As the father of a two year old girl, let me tell you that did not work out well for Anna in Frozen and it probably won’t work out well for you either.

In the old school digital marketing world, there’s the idea of a “tripwire” product where you get a prospect to buy something small (like a $7 ebook) and then on the back end try to sell them into something bigger (like a $1,000 coaching program).

While tripwire products have a bad and scammy reputation to them, they were/are popular because they’re effective. They provide an easy way for someone to get started with you, and you get a customer who is easier to upsell once they’ve bought something.

It is always easier to upsell an existing customer than to get a new customer.

If you’re having problems converting someone to a paying customer, you may be asking too much of them from the start.

Give them an easier way to get started for low cost, and watch your conversions go up.

Remove non-performing keywords and add negative keywords

One of the easiest ways to improve your ROAS on Google Ads is by removing non-performing keywords from your campaigns and keeping your negative keywords up to date as they come up.

On the non-performing keywords part, these are keywords that get visibility and clicks (thus costing you money) but are not converting well. 

Since you should only bid on keywords that you’ve specifically selected because of their relevance to your business, first you should look hard at your landing pages and ad copy to make sure that they match up both in terms of the actual words as well as their level of your marketing funnel. 

Sending mid-funnel keyword traffic to bottom-of-funnel landing pages will never convert well, so you’d be smart to change that landing page to a different conversion to get that prospect into your funnel. If this still doesn’t work, then consider removing the keyword(s) from your campaigns.

On the negative keywords front, this is an ongoing task that needs to be done consistently as Google will keep trying to serve up your ads as much as possible. They do this because of the idea that more visibility leads to overall better conversions, but this is not always true.

I’ve seen instances over and over where a completely unrelated term is being matched to my campaigns because of a language difference, which leads to clicks that will never convert. For example, an ad containing a keyword that means something else in Spanish was triggering our ads, but our offering is completely different in English. 

We added these keywords to our negative keywords and over time, our spend becomes more productive as we’re being shown for fewer unrelated keywords that drive some clicks that cost us money.

This then leads to not only lower ad costs, but also higher conversions from the clicks we do end up paying for. Hello improved ROAS! As a sidenote, this is why “cost per click” is not the right metric to optimize towards, because you might be getting cheap but totally unrelated clicks. 

ROAS is by no means a perfect metric (ROI is better), but it’s much better than cost per click as a KPI.

Time your ads and promos with the right time of year

Holidays and time of year are known to dramatically swing ROAS, and which ones swing it which direction depends on your type of business.

For example, travel companies see dramatically better ROAS during travel booking periods of the year but likely dramatically less right at holidays when people are already or have already traveled.

Ecommerce companies on the other hand see worse ROAS during some months and better ROAS during shopping periods (even though ad costs are higher, sales are much higher as well).

By pulling back spend when people aren’t in as much of a buyer mode, you can improve your advertising performance without leaving much if anything on the table.

Focus your audience

Next up, sometimes improving ROAS can be done by being more specific about the audience you are targeting. 

Say you’re an ecommerce company selling shoes. 

You sell men’s shoes, so that’s one layer of focus. But “men” is quite broad and so your ROAS probably won’t be great. You may get a volume, but so many are going to be unqualified that it will be hard to make your ROAS make sense.

If the shoes you sell are upmarket, it would then make sense to target men in urban locations since men in urban areas are much more likely to shop online and care about their footwear. This will reduce your reach to be sure, but your hypothesis should be that you’ll get the same number (or close to it) of conversions for less spend because those you’re spending to target and get to your site are your ideal customers.

If you want to go even another cut deeper, you can target men in urban locations who make more than $100,000 per year as a household. You should obviously test this, but if you’re selling high ticket (>$100) men’s shoes this is likely your target audience. Your audience shrinks of course, but conversion rates should also go up.

Test these different setups and closely track:

  • Spend
  • Clicks
  • Conversion rate
  • Conversions
  • Profit

From there, make a call about the tradeoffs. Maybe having a bit “worse” ROAS is an ok tradeoff for you because volume makes up for it. 

You won’t know until you get your own numbers.

Reduce abandonment (emails / retargeting / special offers)

Not every site visitor is going to convert on the first click or visit. This is reality that we have to face.

The question though is how many of those will convert in the future if they are reminded and keep returning to your site to look a few times before buying. “The rule of seven” tell us that on average a person needs to see an ad or a message seven times before they buy.

If this is the case, you need to be everywhere your potential customers are in order to remind them that you exist and they thought they needed you at one point.

There are a few tried and true ways these days to do this:

  1. Retargeting advertising on Facebook, Instagram, and display networks like Google for any type of business.
  2. Email reminders to do things that they came to do – buy something, sign up for something (webinar, call, etc).
  3. In ecommerce, sending abandoned cart emails if someone who has previously bought and is signed in (or cookied) leaves a cart with items in it.
  4. Special offers to entice potential customers to buy now, by offering things like free shipping or a percentage off for taking action now.

The thinking behind all of these strategies is simple – your ROAS (and ROI!) should improve when companies do a better job of keeping in front of potential customers and getting them to come back and buy.

We should note that you still need to track all of these metrics, because it’s also possible to go the other way and hurt your ROAS and ROI by using ineffective remarketing channels. As with all of these strategies, tracking your own metrics and looking at ROAS as a holistic thing (dollars made divided by dollars spent) rather than on one channel is the way to go.

Improve average order / cart value

Finally, one of the easiest ways to improve ROAS is by simply selling customers more stuff!


Easier said than done I know, but if you can improve the average cart size or order value then your advertising dollars will inherently go further.

The math is simple:

  • If you spend $1,000 and get 100 orders at $39 each, your ROAS is 3.9.
  • If you spend $1,000 and get 90 orders at $59 each (an improvement of $20 per order), your ROAS is 5.31. You’ve had 10 fewer orders, but made more ($5,300 instead of $3,900) 

One strategy that has become very popular in ecommerce is to use post purchase upsells. These are easy one-click purchases that a customer can add onto their purchase after the fact, without needing to enter their payment information again.

A strategy that Amazon uses, seemingly to great effect because they have it sitewide and Amazon tests everything, is to show “people also buy” products and even bundle them together to make it easy for someone to buy all at once. Here’s an example from a truck hood modification I’ve looked at on Amazon, and how they try to upsell me to other things people buy:

In B2B, you can improve your ROAS by upselling, adding on services, or even extending the average amount of time a customer works with you. This improved customer lifetime value (CLTV) also serves to reduce your ROAS because you are making more over the long term.

A final word about measurement

An area where we most often see companies failing in their marketing is around measurement of their efforts and the return they’re getting.

Attribution is notoriously very difficult to get right. There are many different models (first click, last click, a sliding scale) and none of them are perfect.

But don’t let the desire be perfect get in the way of getting something good enough. Some knowledge is better than none, and perfect doesn’t exist.

As Mark Twain is once purported to have said, “There are three kinds of lies: Lies, Damned Lies, and Statistics”.

Establishing a cadence of measurement and looking at results and trends over long enough of a time horizon will give you a much clearer picture of how well your channels are performing. In light of seasonality, it is imperative that you look at as long of a time horizon as possible in order to avoid making long term decisions off of short term data.