Conversion Analysis – David Meiselman

May 3, 2009June 11, 2009

Turning website visitors into buyers – Online Funnel Part IV

Presenting a product for purchase – moving your prospects from visitor mode to buyer mode.

For those sites that monetize by selling products this is often the most important step of the funnel. It is all well and good to have people read your helpful content and even to continue to engage with your site or newsletters on a regular basis. But while you want to grow an audience, you need at least a chunk of them to eventually convert or else the exercise is not commercial. Likewise, if you are ad supported, you need some level of demonstrable ROI to show your advertisers, unless they are just concerned with brand impressions, which is less and less common. So someone needs to click on something.

How do you get visitors to consider buying something and how do you measure it?

Some people will measure the conversion from relationships to customers as one big step. I prefer to break it down further to two steps (or even three): purchase consideration and store check out process. Each of these steps is measurable and is effected by different drivers. If optimization and revenue maximization is your goal, it helps to attack each step separately.

The first step is to get your relationship or visitor into the store or into the mode to think about buying your product. The key metric here is the visit to the store or product page, though some tactics skip this step all together and drive clicks directly to the shopping cart itself. The key tactic is to present the target with an offer. This can either be in reaction to users, such as displaying a special offer callout on a content page they are visiting, or can be proactive, such as sending the user an offer specific email.

In future posts, we will go into much greater detail on the strategy and tactics behind specific emails and broader communications approaches. There are a wide number of inputs that drive their success, from the frequency of touch, to the mix of content and commerce, to the specific types of offers you present and the creative you use to do so.

For today’s post, we will stay at a macro level and note the fact that the key aspect of conversion analysis here is to measure the effectiveness of tactics based on both the click through rate (CTR) they achieve, as well as the downstream metrics of orders/click and dollars/click. As we have said before, too keen a focus on the immediate metric (CTR) can lose the forrest for the trees. At the end of the day, it is all about the dollars that come from the activity. So do not stop your analysis between approaches at the one that produces the best CTR.

You can actually logically connect certain creative approaches to higher click through rates that have lower sales rates. Pitching something for free that actually has a cost is the clearest example. In so called “direct PPC” campaigns that focus on selling something immediately to those who click, driving down CTR can actually have a highly positive effect by focusing on those who are serious about buying something vs. those who never intended to buy, but were “just looking” or seeking a freebie. When you are paying for that click, you want a high ratio of buyers within clickers.

So if the key metric is visits to the store, which in turn is driven by CTR of site visitors or email readers, etc., what are the core drivers to effect that number? The main variables are within the offer or the pitch itself. What product are you offering? Do your prospects know the product? Where is the price point? Is it discounted? Are there special aspects beyond price, like free shipping? How are you describing the offer? What benefits are you presenting? What copy or visual creative are you using? Each of these elements can be tweaked and tested to optimize results.

Outside of the offer, what other factors drive success? How relevant is the offer to its context? In other words, does it match the content it sat next to? Displaying call-out ads on a site that relate to the content pages themselves will tend to have a much higher rate of engagement. If you run a home improvement site, ads for painting gear will generally drive better results sitting in the painting how-to section than generic home-improvement ads in their same spot. This sounds obvious, but it is amazing how many sites take a one size fits all approach.

In the next installment, we will finally examine the final step of the online conversion analysis funnel, the shopping cart and driving higher close ratios.

Photo by macmuc

April 9, 2009June 11, 2009

Converting first time website visitors – Online Funnel Part III

Leading website visitors down the path to conversion.

They are on your site, now what?

At this point in online conversion analysis, the objectives and their corresponding metrics start to diverge based on what the site does. There are at least as many different conversion goals for site visitors as there are business models or probably many times more. Some product or commerce businesses are immediately focused on a sale. Their intent is to convert that site visitor into a paying customer during that visit or at least during the visitor’s current web session. Others have a long sales cycle and their focus is some indication of interest, like downloading a whitepaper. Still others rely on a longer communication stream to make a sale, so their principal goal is to drive a registration with an email address and permission to talk to the visitor.

Whatever your objective, there is a common need is to create the landing page to serve that goal. It is amazing to me that so many people and businesses still point their paid traffic to their site’s home page. The index page needs to serve a variety of audiences, first time visitors, returning visitors, current customers, press, partners, etc. By definition, there will be a whole bunch of “stuff” on an index page that is of no use to first time users, and often confuses and distracts them from your conversion goal. That does not mean, by the way, that you shouldn’t optimize your index page to convert first time visitors, or even dynamically display different content to first time and return visitors. But a focused landing page will often outperform a more generalist page.

Of course, as we saw in funnel post II, driving traffic to your website comes from a huge variety of sources and programs and each has its own characteristics, objectives, and different types of conversion. Affiliate traffic, for example, might be highly sale focused so it’s primary conversion metric might be adding a product to cart.

Natural search traffic, which often comes from ranking for content within your site, presents multiple challenges and objectives. First, the content needs to serve the reason the user was searching for it. If a user was searching for content on how to sand house clapboards and landed on a product page for a sander, while s/he might be more apt to be interested in the sander than any random web surfer, s/he is likely to move on quickly (or “bounce”) from the page and site if the content doesn’t deliver to the intent. Once the content serves the intent of the visitor, THEN the marketer gets to try and convert the visitor to the next step in the funnel. That goal could be to get the user to click on an ad for a sander or it could be to get the visitor to subscribe to the site newsletter or rss feed.

Other models or other businesses early in a “build a userbase” phase might have very different objectives with first time visitors when compared with commerce focused businesses. Ad-supported content sites might want to convert by tempting the visitor into reading multiple articles. Community based sites might be focused on getting the visitor to create a profile. Unless your business is highly focused on single commerce transactions, focusing on converting a “visitor” into a “relationship” usually pays the greatest return.

Usually focusing a page on one type of conversion is best, but, as in everything in online marketing, you should test your approach for the best overall results. If removing ads from a page and simply focusing on email signup gives the best results of attributable revenue over time per visit, then do that. Again, as with analyzing your traffic driving metrics, you need to be able to follow conversions all the way through the funnel to be able to assign a value to each type of action. How much is a newsletter subscriber worth to you in future revenues? More precisely, how much is a newsletter subscriber who came from a natural search visit to a deep content page worth? How often can you get the visitor to convert to a subscriber? How often can you get someone to click on an ad for a product that aims down a purchase path?

The more precisely you try to measure these things the more difficult it is to achieve with standard analytics packages. We will look at more precisely measuring values over time in a later post, but the core concepts are that you need to measure the value of different segments defined by their SOURCE as well as the value of segments defined by the ACTION that prompted their purchase or otherwise drove their value. Many systems only measure the later, but in determining which programs are profitable and how to approach converting that traffic, tracking the original source of the customer all the way through to deferred revenue is important.

Once you have your measurements properly in place, testing different approaches to conversion is a fun creative exercise. This is what I always refer to as needing both left and right brain skills. You need to be creative enough to visualize and communicate value AND you need to be analytical enough to hear what your customers tell you and weed through the data to see what works. The only sure thing in testing different approaches to conversion is that many of your hypotheses will be wrong and many will not match best practices. What works for most people will not always work for your site visitors. Short bulleted copy works best for most? Yours might very well react better to long narrative copy. The most important thing is to TEST as many variants as you can come up with.

Assuming your model is one that does not do best when jumping right to a sales pitch, the next step in the conversion funnel (and the next post in this series) will take us past converting the first time visitors into a multi-touch relationships and gets into moving relationships into “purchase consideration” mode.

Photo by kierkier

March 17, 2009June 4, 2009

Driving traffic to your website – Online Funnel Part II

The most basic answer to how we get more $ coming out of the bottom of the funnel is to add more people to the top. Driving more traffic is an obvious goal and it is often the focus of online marketing activities seeking the classic hockey stick graph depicting a steep climb in site visits. For online conversion analysis, it is stop number one.

In fact it goes back to sales 101. My first sales manager at Merrill Lynch a thousand years ago (ok, maybe not quite that long) used to say that you know how many people you have to speak with on average to get a prospect. If it takes 20 conversations to get a live prospect, you should be excited each time you get a “no” because you are 5% closer to “yes”. While this was mainly a tactic to help sales people overcome the negative impact of being told “no” all the time, it actually spoke to a more important aspect of sales in general. It takes large volumes of input to get a reasonable output. Back then it meant, “make more calls”. Now it means “drive more traffic”.

There are countless ways to drive traffic to your site or store. Each one is measurable in its own way and each one has its advantages and disadvantages. We will drill into the specifics of each traffic driving tactic in later posts, but a non-exhaustive list in no particular order includes such things as paid search (adwords, etc.), natural search, affiliate programs, social media, direct outreach through email, referred traffic from links, word-of-mouth and viral marketing, etc.

It should also be mentioned that “traffic” is not always measured as visits to a website. In a later post, I will discuss how online interactions and “touchpoints” are increasingly taking place off the site, through such points of interactions as embedded widgets, the reading of feeds, etc.

All website analytics packages give marketers visibility into the source of their traffic and help them track which of their traffic driving programs are yielding the best results for growth in visits. Where measurement becomes trickier is when you start to look at the impact increased traffic has on your downstream metrics. This is key. It is not simply enough to focus on each step in the funnel as a separate and unrelated metric. You cannot assume that new volumes of traffic from new sources or new campaigns will behave and convert as the previous flows did. If you are lucky enough to be in a business where your conversion rates are rock steady regardless of who walks in the door, lucky you. The reality is that different subsegments of visitors will behave very differently at each step in the funnel. Visitors driven by paid search will act differently across the search engines that send them, the keywords in the search, their own geography, etc.

The important lesson here is that even when focusing on step one in the funnel, marketers must follow new traffic flow all the way through the process to determine if the effort or expenditure is worth the cost in time or dollars. Now, don’t get me wrong. There are secondary benefits to site traffic beyond whether a specific visitor ultimately monetizes – growth in your brand recognition, potential referrals through word of mouth, additional links to your site, etc. But if you have limited resources and need a direct ROI on your investment, you can’t simply grow the top of the funnel without following each new segment through the whole process.

At the end of the day, online marketers should be able to calculate the cost of a given visit and match it up with the ultimate revenue over time driven by that visit. It is then that differences in business model have the biggest impacts on whether a given traffic driving campaign makes sense. If you sell stuff and most transactions are one time events, you need to make that cost of acquisition fit in a very defined and limited range. If your model drives recurring revenue (through subscriptions or upgrades, etc.) you can afford to spend more to acquire the first transaction and thus more on driving the original visit.

The big lesson is to measure your cost of traffic driving but consider the needs of your business before you determine if a campaign is worth the traffic it drives.

Of course, many “traffic driving” programs involve the need for content on your site. Whether it is creating “linkbait” for SEO or creating informative content to drive word of mouth or return visits, the user experience is part of what drives the traffic itself. Thus, looking at each step in the funnel is often a fuzzy exercise at best, since they are so intertwined. In the next post in the series, we will look at the initial user experience as step two in the funnel.

Photo by flrnt.

February 8, 2009June 11, 2009

The Online Funnel – Part I

Any website or online business that exists for any purpose beyond self expression has a set of conversions that drive its success. Just like any traditional sales process, online success can be measured as a series of conversions that resemble a funnel. A large number of “targets” or “prospects” is driven into a process of steps that narrow their number down and down until the last remaining targets turn into whatever goal the driver has in mind…customers…users…readers, etc.

This process is often best represented as a funnel, with the large top section getting narrower and narrower with each step, as the pool of targets diminishes with the loss of those who did not “convert” from step to step. Some funnels are short and go right to the ultimate end conversion. Others are taller and multi-stepped, with a more gradual buildup of connection between the user and the site, service, or product.

Why is this important? Well one of the best things about online marketing is the ability to measure success quantitatively and adjust your approach to improve results. When your task is something as large as turning the denizens of the wider internet into customers for your product or service, that is not a single conversion or a single measurement. Rather, most strategies involve many steps and inputs, each of which can be optimized for its greatest contribution to ultimate success. Funnel analysis is a very useful way of breaking down a complex set of drivers into ultimate success and examining the success of each factor. When comparing the metrics for each step to industry standards and to themselves over time or under one approach versus another, online marketers gain a framework for doing their work, applying scarce resources, and seeing where they might have the biggest impact by applying change.

This is the first in a series of posts that will examine the various steps of the online funnel, the metrics that are generally measured for each step, and the tactics that are employed at each step. There are as many different funnels as there are businesses or site publishers on the internet. For our purposes however, we will look at the funnel as consisting of four primary stages prior to ultimate monetization:

1. Traffic driving
2. User engagement
3. Relationship building
4. Monetization events (AKA Purchase decisions)

In its most basic framework, this involves single conversion metrics leading from step to step.

1. Converting measures of traffic to some action on a site or using a service.
2. Moving from a single interaction to a repeated one.
3. Moving from a set of interactions to the consideration of a purchase (for example).
4. Consummating monetization (through a purchase, click-through, or otherwise).

In our next post in the series, we will look at the tactics used in the first funnel step, driving traffic, and the drivers and metrics that lead to conversion from that step to measuring the movement to the second funnel step, user engagement.

Image adapted from original by Normann Copenhagen.