Metrics are used to:
"Clickstream" has been defined as "the database created by the date-stamped and time-stamped, coded/interpreted, button-pushing events enacted by users of interactive media controlling their systems via remote control channel changers, alphanumeric PC keyboards and mice, numeric keyboards of PDAs and similar devices, and voice command of screen media." (CASIE)
Coalition for Advertising Supported Information and Entertainment, an advertising industry group.
Cyberatlas article by Ariel Poler
Internet Measurement: Past & Future
Why do hits continue to be used as a traffic indicator (the recent BusinessWeek cover story, on Net Communities, uses hits as evidence of success at a certain site)? One argument is that they are so easy to get: a site owner doesn't need to spend much in analysis tools or services to know their hit count. Another reason might be that of any measure a site owner can provide, hits are likely to provide largest absolute number, e.g., "we are receiving 5 million hits per month!"
"METRICS AND METHODOLOGY" (9/97)
"The requested elements can be fulfilled from a number of locations:
Internet Advertising Bureau's definition of standard terms
| Term | IAB Definition | Metric | Comment |
|---|---|---|---|
| Ad Request | An opportunity to deliver an advertising element to a Web site visitor. | The request of an advertising element as a direct result of a visitor's action, as recorded by the advertisement server software. | Does not require that the user saw the ad. |
| Click | When a visitor interacts with an advertisement. | The opportunity for a visitor to be transferred to a location by clicking on an advertisement, as recorded by the server. | Does not require that the user arrived at the location. |
| Click Rate | A percentage of response | Clicks divided by ad requests | |
| Page Request | An opportunity for an HTML document to be displayed within a browser window. | .. as a direct result of user's interaction. | Page need not load to completion. Splash pages or interstitials do not count. Frames don't count. |
General Characteristics |
|||
| Browser | Type and version of Web browser | Values are not standardized. | |
| Platform | Type of computer system | ||
| Domain | IP address of user | ||
| Referral | Place where user clicked to get to current page. | ||
User Behavior |
|||
| Visitor | A single user. |
Identify visitor by Unique Registration, Cookie, URL Tagging, IP Address w/ heuristic. | |
| Visit | A series of page requests by a visitor without 30 consecutive minutes of inactivity. | An arbitrary definition, but equivalent to a session. | |
A Sample Log Report
MixNMatch Hitlist report for September.
Historical Data
Aug. 97 Sept. 97 Oct. 97 Nov. 97 Dec. 97 Jan. 98 Feb. 98 Mar 98 Apr 98 Page Views 238,292 281,404 288,967 151,831 148,374 143,656 176,437 192,676 145,400 Unique Vistors 33,752 47,008 65,203 46,571 44,429 41,784 50,728 54,757 45,410 Avg. Pages Per Visit 7 6 4.48 3.26 3.33 3.43 3.49 3.51 3.2 Avg.Daily Visitors 1,205 1,903 1,862 1,663 1,269 1,492 1,811 1,564 1,621
Measure for Measure
There are different tools measuring some of the same information.
Visitor Management Tools |
||
| Tool | Feature | Benefit |
|---|---|---|
| Site analysis | Hit counter | Server load analysis |
| Visit Tracking | Visit length | |
| Path Analysis | Visit patterns | |
| Time-of-day use | Visit patterns | |
| Keywords used | User origination | |
| Referrer Logs | User origination/destination | |
| Ad Server | Cookies | Ad rotation based frequency, platform |
| Time-of-day use | Time-of-day serving | |
| Page Views |
Visit patterns | |
| Impressions | Ad-response profile | |
| Clickthroughs | Ad-response profile | |
| Keyword used | Personalized ad | |
| Referrer Logs | Personalized ad | |
| Personalization | Cookies | Content Profile |
| Pageviews | Content Profile | |
| Click-throughs | Content Profile | |
| Keywords | Content Profile | |
Source: Iconocast, 10/07/98
CPM is the cost to reach a thousand people.
Are Web CPM's too high when compared to other media?
Web ads may do better because of a lower ad/edit ratio.
A typical Web page is 91% editorial and 9% advertising. Dramatic when compared to magazines, which are typically in the 50/50 range, and television, which is closer to 60% programming/40% advertising. (Rick Boyce, IAB, Exploding the Web CPM Myth)
Proctor and Gamble wanted to work with Yahoo to implement an advertising campaign in which what they paid was based on performance.
Most common reason for establishing metrics is to assess the effectiveness of what you are doing.
Source: Iconocast, 9/02/98
Computer industry spends 1.8% of its sales on advertising
Book publishing is 10%
Online businesses might be 14%
"A study commissioned by The Direct Marketing Association and conducted by The WEFA Group, reports that returns on Internet advertising spending has nearly doubled since 1994. Every dollar spent on Internet advertising this year generated $7 in sales, as compared to $4 in 1994. " (Cyberatlas, Dec. 15, 1997)
Advertisers calculate cost-per-click: how much it cost to advertise divided by the number of users who actually click-through to a site.
Click-through initially averaged higher than it does today. Down to 1% to 2%.
Publishers argue that click-through rate fails to account for value of an impression.
User visits site and provides some information to the advertiser
Enter a contest, for instance.
|
Average Marketing cost for each new account |
|
| Company | Cost-per-Sale |
|---|---|
| CDnow | $40 |
| EarthLink | 75 |
| America Online | 93 |
| Credit card companies | 90-100 |
| Long distance telephone companies | 100 |
| Mortgage lenders | 100-250 |
Source: Iconocast, 9/02/98
Ideally, we want to be able to know more about our customers, understand their preferences, learn from their history.
Project 2000 (Novak and Hoffman)