Above the Crowd

Conversion: The Most Important Internet Metric of All (Revisited)

October 2, 2013:

Over 13 years ago, in March of 2000, I wrote a blog post titled “The Most Powerful Internet Metric of All.” The key thesis was this: if an Internet company could obsess about only one metric, it should be conversion. No other metric so holistically captures as many critical aspects of a web site – user design, usability, performance, convenience, ad effectiveness, net promoter score, customer satisfaction – all in a single measurement. Yet despite the remarkable power of this metric, it is alarming how few companies today truly understand conversion and how to optimize it. As such, it is time to pound the table again – conversion is by far the most powerful Internet metric of all.

Internet analytics can be roughly broken down into two simple activities – customer acquisition and customer optimization. Nearly every Internet company on the planet has invested in some form of the first activity, customer acquisition. The most obvious form of this, and by far the easiest to implement, is Google AdWords. As a result, AdWords boasts over one million customers and over $30B in annual revenue. I rarely meet a company that hasn’t experimented with AdWords in some way, and most of those that do understand the basic math around the cost of customer acquisition. Customer acquisition efforts are pervasive in our industry.

There are two reasons why customer acquisition alone is a poor place to establish your company’s competitive advantage. First, the sheer intensity of the competition for effective AdWord inventory reduces the likelihood of a sustainably high ROI. Second, ad spending does not provide leverage. Each year, your company will want to grow by a higher absolute dollar amount than the previous year, and as a result you will need to buy even more traffic. This is the piece that the LTV-zealots always miss: the spending will never stop.

The second activity, which I believe should be the obsession of all analytical startups, is customer optimization. Think of this as understanding how to optimize the performance of the customers you already have. This is the art of conversion: improve your site’s conversion and you simultaneously increase operating leverage AND increase competitive differentiation – a truly powerful combination.

Basic conversion is easy to understand. In its most simplistic form, it is the ratio of customers that achieve some desired action or activity divided by the total number of visitors to your web site. Your web site likely has a single overall conversion rate (the table below highlights one firm’s view of industry-by-industry “overall” conversion rates). However, in order to truly master conversion you have to understand that there is not just one conversion rate, but numerous conversion rates throughout your site. You need to understand the rate for each and every landing page. You need to understand it for each and every source of traffic. You want to understand it in each section of your website. You want to know what it is for each and every outreach campaign. You want to know and understand bounce rates for each and every page. “Conversion Rate” is a misleading term. To truly understand conversion you need a “conversion matrix” for every path through your site.


Even this does not tell the whole story. In 2013, we are in the midst of a critical platform transition from the browser-based Internet to mobile applications platforms like iOS and Android. This transition adds even more complexity to the conversion puzzle, as you now need to track and understand conversion on mobile as well. The bad news is that the tools for doing mobile conversion measurement are immature at best. The good news is that if you can master them at this early stage, you will have a significant competitive advantage.

The best place to focus your efforts to improve conversion are within the product itself. Conversion at its essence is a proxy for usability and convenience. Compare Amazon’s one-click experience with an e-commerce site that requires five pages of information input in order to affect a sale. Which do you think will have a higher conversion? Conversion is also influenced by effective personalization. If a site is designed to respond to my particular needs, I will obviously be more inclined to transact there. This is the dream of “big data” – to harness the information you already have to deliver a personalized experience for each user.

Conversion is also highly correlated with a site’s speed and performance. No one likes to wait around for page loads. Once again, this is something that needs to be measured all over your site and not just in one place. You home page may load plenty fast, but your shopping cart may take way too long to load. You have to attack performance everywhere. There are also many who believe that site speed positively impacts SEO, and as a result, you have multiple reasons to care about this.

Unfortunately, conversion improvements typically are the aggregate gain of 100 tiny improvements, not one silver bullet. Rarely will you find one single change that is going to have a 5% lift in conversion (you might if you have never tried, but this type of win will go away quickly). Rather you will find 30 things on a page that all have a tiny impact, and the overall impact, after months of work might be 5%. You have to be willing to toil in the minutia knowing that the impact on the overall system will be the combined result of many tiny little changes.

Despite the huge potential for gain, conversion improvements are fleeting for many organizations. Many product teams view web-site design as more art than science. Highly paid designers are employed to create beautiful web pages and mobile applications, and their opinions are coddled as divine wisdom. Measuring the efficacy of their divine wisdom at 100 different places in your customer flow funnel can often prove to be controversial, and it can produce friction between the marketing department, who often owns customer optimization, and product development, which owns design. Not all organizations can operate with the proper amount of transparency and self-awareness, or have the patience to try a number of different approaches to produce very small incremental gains. The companies I have seen that are most successful with optimization are assiduous in their approach.

So, given how hard it is to improve conversion, why bother? Because improving conversion, even a little bit, can produce insane economic leverage for your business. The table inserted below looks at five leading consumer Internet companies – Blue Nile, TheKnot, C-Trip, Trip Advisor, and Priceline – and models what would happen if they improved conversion 10%. Just to make this clear, we are talking about moving from say a 6% overall conversion rate to 6.6%, or simply a 60 basis point increase in absolute conversion. Because conversion improvements impact the customer flow you already have, they lead to a direct increase in revenue, offset solely by the cost of running the conversion improvement effort. In this case we used $1mm/year cost for the smaller companies, and up to $10mm/year for Priceline.

excel table

If you look across the group you will see impressive increases in operating income, and net income margin percentage. Additionally, the ROI on the conversion costs range from 650% to 4000%! Even if you only achieved a 5% increase (versus the 10% we modeled), you would still see ROI in the 275%-1940% range. And for a 6% converting site, we are once again talking about a 30 basis point improvement in absolute conversion rate. Tiny moves have huge financial consequences which almost seem magical or unfair. You can build a similar model for your own business, and put in the inputs as you see fit, but you will not find another single area in your business where small improvements will have such a powerful impact.

Furthermore, the larger your company, the more you should increase your focus on conversion. With a large customer base, you are able to leverage any improvement against a much broader audience. You can see this in the model – bigger companies get a bigger advantage, even when we estimate a 10X larger budget for conversion improvement. Investing in a conversion improvement effort is a no brainer for these larger companies.

All companies should prioritize customer optimization ahead of scaling customer acquisition. Think about customer acquisition as buying fuel, and customer optimization as measuring the fuel efficiency of your engine. Ironically, many more companies focus on buying fuel in volume than they do improving fuel efficiency. The whole world has it ass-backwards. You should focus on fuel efficiency first, and when you know that it is solid, then you can start buying fuel in volume. Yet 90% of the companies I meet are buying and buying customers (acquisition) without any sophistication around understanding conversion (optimization).

Unless you are fairly certain you have a high conversion rate versus your peer group, you can be at an extreme disadvantage buying high volume advertising. Conversion rate directly impacts how much you can affordably spend on advertising. If you have 2X the conversion rate on your competitor, than you can spend twice as much and get the same outcome. Likewise, if you have a poor conversion rate versus your competition, than you are bringing a knife to a gun-fight. Ever been a room where an entire team all agrees that your competitor must be an idiot because they are spending so much at such high rates on Google? The answer may be that they have a much higher conversion rate, and that you are eligible for the role of the idiot. Don’t push the gas pedal if you don’t know your fuel efficiency. Likewise, if you know you have a conversion advantage, you also have an ad buying advantage.

Benchmark has such conviction about the amazing power of conversion that we have made two investments in the past year in companies whose products help drive material conversion lift for their customers.


The first company, Optimizely, has the market’s leading product for A/B testing, the fundamental technique used in improving customer optimization. Optimizely’s impressive SAAS service allows marketers to A/B test using a simple WYSIWYG editor on any page in your site, with zero impact on your engineering team or resources. It’s such a strong value proposition that they have over 4,500 happy customers, including CBS, Disney, Starbucks, and many more. Optimizely’s product should be the fundamental building block for anyone that has any interest in conversion optimization. The product is designed with a clear understanding that optimization is about experimentation, and it greatly lowers the cost and complexity of that experimentation.


The second company, Sailthru, uses its unique personalization technology to help their customers improve conversion. More specifically, clients aggregate all of their customer data points under one roof to provide truly personalized experiences across all platforms and devices (specifically including mobile). Using Sailthru, companies build 360-degree profiles of each and every user that can then be used to personalize emails, site and in-app experiences, search results, and more. Online publishers have seen 10-35% improvements in click through rates that regularly yield 30+% lifts in page views, resulting in substantial incremental ad revenue and profitability. Commerce players have driven 6-10% lifts in average order values, 15%+ improvements in purchase frequency and major improvements in overall customer retention. Most importantly, Sailthru makes personalization achievable and scalable; a near-term reality versus a multi-year project that will never see the light of day.

Looking forward, you should expect customer optimization and conversion analysis to move to the forefront of analytical marketing work. We are likely moving into an “age of enlightenment” where the scientific method is available to all, and ideas win (and are tested) on their empirical results. With improved tools and techniques, and innovative products like Optimizely and Sailthru, highly optimized sites will redefine the new fitness level required to stay afloat in online businesses. Darwin will take care of the rest.

Other resources:

  1. Try out Optimizely today
  2. Video: Learn more about Sailthru
  3. 71 Things to A/B Test
  4. 53 Ways To Increase Conversion Rate
  5. 40 Checkout Page Strategies to Improve Conversion Rates
  6. 100 Conversion Optimization Case Studies
  7. A “Meta-list” on Startup Management (Thx:William Mougayar)
  8. (send me more and I will add them to the list)


  1. Drew Meyers October 2, 2013

    Awesome analysis, Bill. I don’t come from a conversion background, as I largely did community building/management for Zillow and am more a marketing guy than analytics guy…but I agree whole-heartedly with its importance. I worked for an internet marketing agency (Virtual Results) for a year after Zillow, whose entire business was based on testing every last component on the sites we built to optimize for converting traffic into leads for agents/brokers. So it brought me more into that camp…which of course I’m thankful for 🙂

  2. tigger October 2, 2013


    I agree wholeheartedly with your post. Based on my experience at several startups, there is something more fundamental going on with companies that don’t track conversion. It’s not that they don’t get the logic of what you are saying. Rather, they fundamentally choose a “creative” approach to product.

    In early companies, the first iteration is a creative endeavor … usually solving a problem that they intuitively know about. At some point, I believe it becomes appropriate to measure and improve through sample-based iteration. Unfortunately, many product founders aren’t willing to shift gears and listen to their numbers (i.e., and sometimes they won’t even listen to customer anecdotes). Rather, they “know best” and continue their current focus.

    This culture war betweens the “creative” vs. “analyticals” is not restricted to small companies. As a gross oversimplification, consider Google and Marissa Mayer’s testing of 30 shades of blue on the homepage. While I, as an analytical type, love the story. Many, many engineers and designers that I’ve met hate it and everything it represents. As an outsider, I sometimes wonder if much of Google product isn’t run by folks that guess rather than measure what people want to do … brilliant engineers but…

    Finally, I wonder if this cultural divide isn’t about control. After all, if I have to listen to my customers and the numbers, then who is the boss? Maybe, it forces me to work on product that I don’t find interesting.

    At some point, you have to decide whether you’re running an art show or a business. You know where I stand.

    PS: My apologies if this is too tangental.

    • bgurley October 2, 2013

      Actually – I think you are spot on.

    • Edwin Khodabakchian October 6, 2013

      You actually need both to be successful: you need to design like you are right and listen like you are wrong.

  3. William Mougayar October 2, 2013

    Hi Bill,
    Great insights as usual. Conversion Rate Optimization (CRO) is the new retail merchandizing, topped with much better data than retail merchandizing had. You got them inside your store (website), now how do you maximize their actions?

    Only thing with A/B Testing, is the permutations can get mind-boggling, and you end-up doing too many of them to figure out that only 10-15% mattered (from an actionable perspective).

    So, for active/high-volume websites, you need a front-end tool prior to A/B testing that can identify the best elements to focus on, and that can get you up to 30-40% hit rate on optimization effectiveness. There’s a company I recently came across that’s focusing on that part, Pathful http://www.pathful.com. So, the trick is to test what matters, and that’s not always so obvious.

  4. William Mougayar October 2, 2013

    As for links on Conversions, check out this curated page on Startup Management specifically (it includes most of the above links you have):

    • bgurley October 2, 2013


  5. Amanda Besemer October 2, 2013

    Awareness without action is meaningless… conversion is marketing 101. It’s not novel, and it’s not rocket science, but it can be perfected via marketing science. Somewhere along the line with the internet we forgot this. New fangled metrics in “Geek Speak” tripped up even the most sophisticated “old school” marketers. I say bring in the old school marketing – the rules have not changed, simply the form. Good post.

    • William Mougayar October 2, 2013

      I’m with you on the trip-up which is to assume that marketing starts with conversion, whereas it doesn’t. It starts with awareness and discovery which then leads to giving a chance at conversions. That said, marketers still have to understand the new science of online conversion analytics. It’s like online merchandizing.

  6. Saul Lieberman October 3, 2013

    Works nicely with Seth Godin’s post (also yesterday):
    Coordinate and amplify
    If you’ve got an idea or you’re working in marketing, the temptation is to seek out and evangelize those that ‘don’t get it,’ to find and sell to the skeptics.
    In fact, real change comes from finding and embracing and connecting and amplifying those that are inclined to like you and believe in you.
    Ideas spread from person to person, not so much from you to them. So find your biggest fans and give them a story to tell.

  7. Sankalp Agarwal October 3, 2013

    I would consider the argument of conversion % of lift of 5% etc.

    I get the gist but I would like to mention something amazing about conversions, which you have missed.

    “Customers respond to value adds in ways not imagined”

    Our conversion rate was 50% more than industry standards by A/B testing. This was achieved by adding a lot of small small items finally adding up the maths.

    But, later I saw straight 2X improvements in conversions by offering customers value that they might want now and not directly offering my product.

    While everyone is busy in doing A/B testing, they miss out on adding value to customers in more ways the customer can thank you for. !

  8. Amit Somani October 4, 2013

    Great article. While conversion is key, the most important metric is revenue per user. This ensures you can’t discount your way to increasing conversion. Further, if you want to take effectiveness of marketing into account, you could compute net direct contribution per user. This (trended over time) can give you effectiveness of both customer acquisition cost and of your conversion machine.

  9. Roger Pavane October 4, 2013

    Great points are made here. The one fundamental problem is budget. In order to do this right you need to pour money into the initiative. As a marketing, business development and new media profession in the space since the beginning of time old school / new school it still costs. Would love a discussion surrounding this topic specific to spend and budget.

  10. Ashish Soni October 6, 2013

    Great post Bill!
    SEMOZ has this amazing 91 step e-commerce conversion map/detailed list.

  11. Josh October 8, 2013

    Lovely article, Bill, on conversion rate optimization. What you presented was a very strong and compelling case to shift more of the focus on a website’s conversion rate. I even liked the companies you mentioned at the bottom of the article, but want to make sure you get a chance to see the next wave of conversion solutions that are about to make a big impact in the world of e-commerce. Moving from simple A/B testing (and making adjustments the old fashioned way), to automatically optimizing A/B tests is a giant leap forward.

  12. Chris Goward October 8, 2013

    I couldn’t agree with you more, Bill! That’s why I started WiderFunnel in 2007 to focus on finding the best methods to radically lift conversion rates and revenue for companies in all industries.

    Here are 30+ case studies showing some of the results so far: http://www.widerfunnel.com/case-studies

    And, here’s where we share the insights from all of this A/B testing research: http://www.widerfunnel.com/blog

  13. Avinash October 22, 2013

    Hey Bill,

    It was really great reading the article. Just to conclude i want to know if you are asked to give percentage division in business to conversion and Landing page/web page or apps, how much will you allocate. I do agree conversion is something really really important but the surrounding factors give a much heavier impact on catching a user. So just to know what will the ratio you will assign with respect to conversion metric with other factors?

    Asst. Manager-Adops – xAD

  14. AJAY PRAKASH October 22, 2013

    It’s a good read but problem with digital advertising is that we can measure little too much. Advertising isn’t only about tracking $/ INR spent, but about building brands.

    Whenever I asked the question “ why digital is better than usual media” to the sales / agencies guys, the maximum answers were on the lines that it helps to track ROI with metrics like CPC, CPL, CPD,CPI and what not. We over complicate the medium with complex metrics. This positivity has become the biggest weaknesses of digital medium. Marketers run after ROI measurement…. through secondary clicks received or downloads etc. and don’t care a damn about other important parameters like brand building.

    That could be the possible reason why we don’t have many good examples of brands built entirely using digital (mobile) advertising.

    I am sure with all the technical expertise, digital advertisement can also measure some other metrics like Brand Recall, Brand Recognition etc. which can give TV/ Print media run for their money. Customer acquistion and conversion are two important metrics but they have overshadowed other equally or more important metrics. I think whole industry sold digital in a very wrong way.

  15. Jennifer Bailey November 7, 2013

    Agree that conversion is critical but is even more complex than described due to the increasing rise of customers shopping across multiple devices. For customers researching on a phone, but then purchasing on the web or iPAD, cross device/platform conversion is the next right step in understanding true conversion. Not purchasing on a mobile platform is not necessarily a usability optimization issue (although it often can be), but is also often related to the on the go nature of phones (do I really want to focus on buying an item when I’m rushing to my next meeting ?). With more and more traffic coming from mobile devices, web conversion will see downward pressure and understanding the cross platform behavior and conversion of users both in web, mobile web and in apps is critical.

  16. Chris January 2, 2014

    Great article! I find it interesting that Google doesn’t spend more money on increasing conversions for their customers. I know they have google analytics, website optimizer, and a number of other tools, but I would think they would benefit from investing in this even more heavily.

    The simple math for my company is such:

    Profit per Conversion = Conversion Value – (Cost Per Click) * (1 / Conversion Rate)

    I can bid a higher CPC if I can increase my conversion rate or increase my conversion value. Google can directly influence one of these metrics….

    Now just because I can bid more won’t mean I will, but since I am a price taker in this instance and Google sells to a fragmented bunch of customers with no power in this position, I will ultimately ending up bidding more once my competitors figure out how to increase their conversion rates. Anyway, it seems like it would greatly benefit Google to invest more resources into increasing conversion rates for their customers.

    • bgurley January 24, 2014

      Great point.

  17. Chris January 24, 2014

    Thanks. FYI, we have a VCs vs. Entrepreneurs basketball pickup game in Boston every Tuesday morning. If you are in town you should play! It’s very informal but we have a great time. Let me know if you are interested 🙂


Leave a Comment

Your email address will not be published. Required fields are marked *

Prove that you're a human *