Improving Web ROI: SEO / eMarketing cost comparison to direct mail

As online commerce becomes ever more pervasive and the Web becomes just another channel that is held to the same standards as any other, there is increasing focus on getting better ROI from technology investments. Vendors in fields from search engine marketing to usability to Web analytics are feeling the pressure to justify their products by claiming they can improve ROI, and the trade press has shifted from a discussion of soft benefits to calculations of how Web technologies impact the bottom line. This is a welcome shift, but because many discussions of Web ROI are so focused around a single sub-topic and so clearly intended to present a vendor’s solution in a favorable light, many business managers are understandably skeptical of any discussion of Web ROI. Decisions to invest in technology --- particularly the decision to move into a new area such as search engine marketing or pay per click advertising--- are still often made more on instinct than on firm calculations, and it is often only after gaining sometimes painful experience that managers can begin to understand the real ROI impact of their decisions.

This article will attempt to demystify Web ROI by presenting a clear framework for decision making using a straightforward example case. While we will attempt to present the example clearly, we have not chosen a case where the decision to invest in technology is an obvious slam dunk. Instead, we have chosen the case of a small business that currently has no online sales and marketing capability but does have a great deal of success with sophisticated use of direct mail. This company is a good case to evaluate for a number of reasons:

  • As a small business with fewer than 50 employees, their needs are in many ways as complex as larger companies’, but they do no have the same budget.
  • Because they currently have no online sales capability, our evaluation has to consider the total technology investment required to get started selling and/or marketing online plus the expense of promoting the site through advertising, search optimization or pay per click.
  • Because this company currently has success with a very clever use of direct mail, the ROI bar for an investment in marketing or selling online is particularly high.

Before we evaluate the case for an investment in selling and marketing online for this company, we will first look at the company background --- what business they are in and what they sell. We will then look at their current marketing efforts focused at direct mail. By understanding the ROI of these efforts, we can get benchmark against which to evaluate online sales investments offline.

Company Background

The company we’ll evaluate is a multi-site auto insurance agency specializing in high risk coverage. They currently have an informational website, and some of their back-end services to partner agencies are provided via the Web, but they do not have online e-commerce capabilities. Through commercially-available software, they have the ability to generate comparative quotes for multiple insurance companies, and they could move to online quoting with only a moderate investment in e-commerce capabilities. The big unknown, then, is ROI --- whether sufficient sales could be generated online to compete with their current direct-mail dominated marketing.

Direct Mail Success

In many ways, this company is in the best possible position for direct mail success because they can acquire free lists of people who may be prime consumers of services like those they provide. The company has access to public records data of drivers convicted of violations that put them in the high-risk category, and those drivers are legally required to maintain liability insurance in order to keep driving.

This firm sends monthly mailings to newly identified high-risk drivers explaining the benefits of the agency’s high-risk insurance products, and they gain new customers when recipients call the included phone number or walk into one of the agencies locations. Mailings average 15,000 pieces per month. Costs for each monthly mailing are $3000 in postage, $1000 in printing and paper, and roughly $150 of staff time. They spend approximately $5,000 per year on direct mail sorting and address verification software and services, or $417 dollars per month. This brings the total cost for each 15,000 mails sent to $4,567.

The response rate for direct mail is often calculated based on how many customer contacts a mailing generates, but a more useful metric (and the one this company uses) is the number of sales that each mailing generates. On average, each 15,000 mails generate 75 new customer insurance policies, meaning that the cost per new customer acquired is slightly more than $60.00. This represents an excellent ROI for the company, because not only is the initial cost more than covered by commission on the initial policy, but customers are likely to renew, generating additional commission at a very low cost.

Technology Investment for Search Engine Marketing

In order to evaluate whether selling online would make sense for this kind of business, we need to reach an apples-to-apples comparison with the cost per policy sold we calculated above. To get to this cost, we need to first understand the conversion rate we could expect to achieve on the agency’s website. The conversion rate is the percentage of visitors to the site who are converted to customers --- if out of every 100 customers who come to the site, 2 eventually become customers, we would have a 2% conversion rate. Conversion rates vary wildly by site, but most experts estimate the average to be between 1 and 2%.

Since the company has an informational site with no e-commerce capabilities, to begin trying to acquire new customers online, the first decision they would need to make would be what part of the sales process would be online, and this decision would have substantial cost and conversion implications. In general, the more convenient the online purchase process and/or the more full service the site, the higher conversion rates will be.

Our example company could, for example, choose to use the site as a basic lead generation tool. For a minimal investment (let’s call it option A), they could update their website to provide detailed information about their insurance products and allow customers to request a quote by submitting demographic information and data about their driving record and cars through a basic Web form. Customer service reps would then generate a quote offline and respond to the customer by phone. This option would involve the smallest technology investment, but it would have by far the lowest conversion rate. With this model, we could probably count on a conversion rate no higher than 0.5% and a technology spend of at most $5000. It is also important to remember that by not automating more of the process through the site, the company’s customer service people would still have to generate quotes and do other administrative work manually.

The next step up that we might choose in terms of automating the purchase process (option B) would be basic online quoting so that the customer could see approximately what the policy would cost. This would require more investment, but it would provide a higher conversion rate because customers could complete more of their decision process (if not the actual purchase) online. Using this model, we might be able to achieve conversion rates of 1%, but we might spend as much as $20,000 on technology. Since we were automating the quoting process, we might also save $2 per policy on the back end in customer service rep time.

Finally, we could develop a more complete solution --- offering comparative quotes from several insurers or even allowing customers to purchase insurance online (option C). Using this approach, we might be able to generate conversion rates of 4% based on convenience, and we could probably automate $4 worth of processing per policy, but this more advanced solution might cost $50,000.

Note that these estimates of number of customers, conversion rate and cost are rough. An important part of our project would be to research and determine how many customers we think we could attract, what kinds of conversion rates similar sites achieve and how much the technology infrastructure would cost. The point here is to understand where we need to focus our research and, once we have estimates, how to calculate ROI. This will allow us to build best, worst and likely case scenarios and make decisions based on those.

Online Promotion

In addition to the initial investment in technology, we would need to spend money to promote the site, and here we have basically three options:

  • Pay per click advertising. This involves contracting with a search engine to display “sponsored” or paid links to a site along with the organic results from the search engine for a given set of keywords, and for the engine to charge the site owner each time a user actually clicks one of these links. Links are displayed based on advertisers’ decisions about which keywords they would like their ad presented for (e.g. “high risk auto insurance” in our case), and in order based on advertiser’s bids for position --- a higher bid means that the advertiser’s ad appears nearer the top. Pay per click is widely considered an excellent advertising value because results are targeted to people looking for terms related to the advertiser’s products, and the advertiser only pays when a user actually clicks through to the advertiser’s site.
  • Search engine optimization. This involves structuring the informational content of the site in such a way that it will be attractive to search engines. By trying to build pages so that search engines determine they are relevant to terms customers are interested in (e.g. “high risk auto insurance”), we can try to make our pages come up near the top of the “organic” or non-paid results when someone searches on these terms. Ranking near the top of organic results is like a windfall for companies that can manage it, since engines do not charge for displaying your link in these results, but it does take effort and usually investment.

Offline promotion. This involves simply promoting your business in traditional ways, but in addition to promoting your phone number, promoting your URL.

Calculating ROI

In calculating the ROI of a move to online selling and promotion, we need to compare how our total technology and promotion cost per sale stacks up against the almost $69 per customer calculated above for direct mail. To do so, we need to make some assumptions:

  • We assume the payback period on our technology investment is about two years. We may not be using the exact same technology in two years, but by being better positioned for any technical upgrades in the future, we will get value from an investment now for at least that long.
  • We assume some number of customers that we could realistically get each month based on the size of the market. We will use this number to calculate how much we can afford to pay for each click to our site in order to break even. We could also assume how much we will pay per click and calculate a break-even number of customers --- either approach is valid. The number of customers we are able to serve is often a key difference between online and offline channels --- if we can create a project that has a positive ROI per click based on a conservative number of customers, then we have a real opportunity if we can attract more clicks and conversions, even if we still pay our break even rate per click. This is because our fixed cost for technology infrastructure will be spread out over more customers. Since we are automating much of the selling process, we are not as constrained in increasing sales volume as we would be using offline channels. We will initially assume the same number of customers that we get from our offline efforts --- 75 per month.
  • In order to calculate how much we will save or lose, we must make an assumption about how much it will cost to get people to our site ---- some “cost per click.”

Based on these assumptions, the ROI for our investment in technology is shown in the table below. The key to these calculations is the conversion rate, which tells us how many clicks we need to get in order to make a sale --- 200 for option A which had a 0.5% conversion rate, 100 for option B and 25 for option C.

We also calculate a technology cost per customer by dividing our monthly technology cost by the number of new customers per month. This monthly cost, minus the savings from automation per customer, gives us a net cost per customer. Subtracting the net cost online per customer from the cost per offline customer gives us a gross savings per customer. This amount is what we can use in online promotion (search engine optimization and/or pay per click) and still break even.

Since the cost of driving traffic to our site is best calculated as a per click cost rather than per sale, we need to understand what we can afford to pay for each click-through to our site, which directly relates to our conversion rate. Since under option A with a 0.5% conversion rate it takes 200 clicks to get a sale, we can only afford to pay $0.29 per click ($57.22 savings per customer/200 clicks to get a customer). As our technology cost goes up under options B and C, so does our conversion rate, so we can afford to pay more per click and still break even ($0.51 and $1.45 respectively).

    Option A

    Option B

    Option C


    Cost of offline customer:





    Estimated # of new customers per month:





    Technology payback (months):





    Technology infrastructure cost:





    Conversion rate:





    Clicks to get a sale (1/E):





    Total clicks we need per month
    (F x B):





    Technology cost per customer
    (D / B):





    Customer service savings per customer:




    Net technology infrastructure cost per customer (H - I):





    Gross savings per customer versus offline (A - J):





    Break even cost we can pay per click in promotion (K / F):





    Estimated actual cost per click:





    Monthly savings from selling online ((L - M) * G):




We can see that while it is the most expensive to set up, option C will have the highest ROI --- if our assumptions about conversion rate and actual cost per click hold up. This example makes clear the key role of conversion rate in determining our ROI --- because our conversion rate is eight times better in example C, we only need 1875 clicks to get our target number of new customers, and we can better afford our assumed cost per click of $0.50. If, on the other hand, our conversion rate for option C were only 2%, we would only save $841.67 per month, which is not as good but still much better than $66.67.

If we play with assumptions about the volume of customers we could get, we can see a dramatic impact. At 25 new customers per month, our ROI looks like this:




    Savings per month:




    And at 150 customers per month, our savings per month looks like this:




    Savings per month:





We can see from these scenarios that the key to understanding the ROI of a Web project is being able to estimate several key factors. You must have some sense of the number of customers that you can attract, which is a function of both the number of potential customers out there for your business and the amount of competition. In our example, the insurance agency will currently be limited to service customers in the state in which it is licensed. This limits the market size, but it also limits the amount of competition the company could expect because they can focus their promotion efforts at their local market. Several tools exist to help with estimating this target market and the costs of attracting customers, and iData can assist you with applying these tools to your particular situation.

You must be able to estimate the number of conversions that you can expect. As you plan a project, you must understand how to measure conversions and how to evaluate your site for usability issues that can impact conversion. iData has usability guides and checklists, and our Visitrak TM analytics product allows for precise tracking of traffic and conversions as well as several other key metrics.

Finally, you must be able to estimate cost per click to get visitors to our site. An easy way to get started with a site promotion and tracking program is to begin a small scale pay-per-click campaign using Google’s Adwords TM and/or Overture’s Precision Match TM pay per click programs to begin to drive traffic to the site. Using a pay per click program in conjunction with Web analytics tool such as Google AnalyticsTM allows you to get started immediately measuring the cost per click of attracting traffic as well as your conversion rate.