A Web plan should include realistic estimates of projected Web traffic. The first step is to look at the details on measuring projected traffic, and how and where to develop realistic estimates. Regardless of whether your website strategy involves selling products, or bringing in community members, or even developing content, in all cases measuring website traffic and user sessions is a good way for you to gauge how many people are using your hybrid site, and where you need to improve your website in order to get better conversion rates.
Establishing what units to measure
Your first step is to establish measurement units. Measuring Internet traffic is a new challenge. Technology and standards are changing fast. If you ever hear anyone talking about how many “hits” they get on their website you know that they are misinformed. A given page view can produce as many as 100 hits, depending on how many images you display on your pages, rendering the statistics unit “hits” completely useless.
One way the Internet industry measures usage focuses on page views and users. You may also want to measure your banner impressions, as a percentage of total page views, if you are hosting advertising banners.
Two favorite measurements are page views and unique user sessions. If you can track the number of unique users that visit your site, and know how many of your website pages the average user views, then you have some very valuable data. You can tell potential investors that you have X number of page views a month, and X number of unique users a month. You can then impress them further by showing the value of your users, and say that each one of your unique users looks at an average of X pages whenever they visit your website.
A page view is a workable alternative measurement in many systems. Each page is a measurable unit. But remember, until you add user measurements to your page views you do not really get a good idea of what is happening on your website.
The math of the traffic forecast can be very simple, depending on the traffic measurements you have available and how you want to forecast. Sometimes you need to add simple math to divide, say, sessions by page views to calculate page views per session. This depends on your forecasting choices. Most often, although the content takes work, the design of the table is simple. It’s built on common sense and reasonable guesses, without statistical analysis, mathematical techniques, or any past data.
The traffic forecast shown here is from a sample Web plan for a consulting company. It includes not just the more easily known information items, sessions and page views, but also the harder-to-derive key ratios that relate the traffic forecast back to the business.
The link between traffic and business is called “leads/calls” in this forecast. According to this projection, the consulting company website will generate one lead/call for every 1,000 page views, (which is the 0.10% number shown). That ratio allows the forecast to apply an assumed closes per leads rate (30% in this case) to derive a projected number of consulting engagements resulting from the consulting website.
As you develop your own forecast, look for ways to relate the traffic forecast to the sales forecast.
What’s a realistic forecast?
The obvious question comes up frequently: how can you forecast Web traffic for a new website, unlike any other, without access to other websites’ traffic statistics? This is a very important question. Here are some recommended answers:
- Ask your Web hosting vendor: Sometimes the best options are also the simplest and most obvious. The vast majority of business websites will be “hosted” by an Internet hosting service. Your account manager at that hosting service should know what his or her accounts are getting as traffic, and what’s realistic. Ask detailed questions, and if the account manager can’t answer them ask for somebody else. Every legitimate hosting service ought to be able to give you an estimate‚not a promise, not a certainty—of what your traffic expectations might be. Ask what other customers of this server are getting in traffic terms such as user sessions, page views, or others? This is a good reminder about asking the vendor to provide statistics as well, on your site, as you ask about other sites? What are the most successful of their clients getting as traffic, and what are new clients getting? What determines the difference?
- Relate your estimate to your marketing: As discussed earlier, your website traffic actually depends almost entirely on what you do to market it to potential users. As you develop your marketing programs, such as searcher strategies and banner advertising, or other programs, try to develop forecasts for traffic. Information is available on industry average performance for banner ads, and links from main searcher sites. Compare the traffic you’re projecting to what the industry experts suggest you’ll get from your Internet marketing.
- Ask experts: If you have access to someone who has been involved in Internet marketing or site development, ask them what traffic estimates might be realistic for your site.
Beyond that, you need to explain the forecast itself. What level of online traffic are you projecting? How fast will traffic grow? What are the most important components of traffic measurement and performance? Why? What traffic measurement will you use and why?
Emphasize important points and explain assumptions. What growth rates are you expecting for the more important pieces in your site, and why? Why are you projecting your traffic at this level? Why not less or more? What are the main driving forces behind the traffic forecast? How does it relate to your market analysis, your main target segments, your sales strategy and marketing strategy? Is your traffic forecast believable? Why?
What risks are involved? What events might turn the traffic forecast downward? What things are you assuming will happen to make sure the traffic goals are met?