7: Monetizing Site Behaviors
Ultimately, every action undertaken by a company is driven by the motive of profitability: to spend less, to make more, and to impact the bottom-line profit. Any expenditure is undertaken with the expectation that it will generate more profit than it costs to implement.
The difficulty in getting support for Web analytics is in its lack of connectedness to the financial metrics that drive an organization. Analysts speak in abstract, statistical terms that traditional business does not understand. To get them excited, and to get their support, analytics must speak in terms of money: "improving the conversion ratio by ten percent" is just noise; "improving sales by $50,000 per month" strikes a chord.
The author provides a hypothetical example of a firm that hires a Web analyst and creates a couple of fictional dialogues between the analyst and the sales director. It's a bit silly and verbose, but basically suggests the value of speaking in terms that others can understand and demonstrating benefit in terms of cost and revenue.
The method for monetizing analytics depends largely on the purpose of a site. For an e-commerce site, it is fairly obvious (revenue earned); for a customer-service site, it's not much more difficult (savings compared to service through other channels); but for a branding site, the behavior of the user on the site is far removed from the behavior that has a financial impact (a sale, made weeks or months later).
In many instances, monetization requires postulation and estimation, and it is unlikely in most instances to be accurate to the penny. However, a high level of precision is not required - such a standard is not applied to any other business endeavor (you do not know the exact impact of an advertising campaign). You merely need to make a well-founded guess, backed by plausible logic.
(EN: The author overlooks a very important element - organizational politics. There is no bulletproof proposal that cannot be shot down by an individual who is looking for an excuse to do so. And often, the objective accuracy of an estimate is less important than the subjective notion of acceptance. That's not to say you don't need some semblance of logic or accuracy, only that it is not always - or even often - the determining factor.)
Sites that generate direct revenue through online sales are fairly simple to monetize: the revenue generated is in dollars. You can also compare the cost of online sales versus other channels and represent this as a savings. It's also been found that customers who use multiple channels are more frequent and higher volume than those who do not, so the Web deserves credit for the increased revenue from the same customers in other channels.
Where lead generation is involved, you can calculate based on the raw number of leads, but also look into the conversion rate (if Web leads convert more often than other leads) and the amount of sales they actually generate. Also, the customer who is a lead becomes a sale and a regular customer with a lifetime value to the organization.
Customer service often calculates the cost per touch and can demonstrate cost-savings against more costly channels. Self-service on the Web is cheaper than phone support or even e-mail support. You should also be able to track customer behavior to determine if those who use online support are more profitable customers in terms of lifetime value (longevity and volume of sales) in all media.
If your site makes referrals, handing visitors off to other companies, you should seek to measure the value of those referrals to others. In addition to the number that click out to other sites, what percentage make a purchase, and how valuable do they become as customers of other firms. And how much more can you charge for advertising if these figures are higher than other sites?
It's also noted that a common behavior is for individuals to gather product information online, but make the actual purchase in other media (an example form a consulting job was that a client found 3% of people purchase online, but 30% of people who purchase in stores use the Web site to conduct research before visiting the store). Aside of surveying, you can attempt to track this behavior by using web-exclusive offer codes or coupons and using a different toll-free number on the Web site than in any other medium.
Even sites that have no immediate impact on income and expense can be said to have a monetary impact. For example, if a site contributes to "customer satisfaction" or "brand awareness," this can be indirectly tied to revenue. If a 'satisfied" customer creates more revenue and the Web site increases customer satisfaction, it can claim credit for revenue attributed to the increase in satisfaction of site visitors. (EN: And at this point, the link becomes even more tenuous and harder to defend based on logic alone.)
Assembling Your Monetization Model
The author refers back to the original hypothetical conversations, to show how the fictional Web analyst derived the dollar-value of implementing improvements in Web analytics. In brief, it's done by determining all the ways that a change in user behavior can impact the profitability of the firm, such that you can make a plausible case for increasing profit by $X based on the differences.
There are a few other details here, basic accounting concepts such as calculating the net present value of making a change when the money spent now will result in savings (and possibly a stream of savings) in future.
(EN: The author continues with a rather extended example of a fictional site an company, which is a good illustration for educational purposes, but merely reiterates the concepts previously discussed, so I'm not preserving notes.)