Is PageRank Enough? PR and Traffic Comparison of 122 Dating Sites

Google uses PageRank (PR) to measure the relative importance of web pages.

PR is useful for ordering search results and measuring the importance of content-based pages but…

Can PageRank be used to evaluate dating sites?

Do dating sites with higher PR offer more value? Does higher PR mean better chances of finding a date?

Let’s compare 122 dating sites by PageRank and traffic to find out if there is any correlation between PR and value to consumers in the online dating industry.

Updated on 2006/11/28: Added Charts


PageRank Basics

PageRank (hereafter “PR”) is a patented link analysis technology that Google uses to measure the relative importance of web pages. At its heart, it’s a vote (by all other pages on the web) about how important a page is.

According to the original Google paper:

“We have created maps containing as many as 518 million of these hyperlinks… These maps allow rapid calculation of a web page’s “PageRank”, an objective measure of its citation importance that corresponds well with people’s subjective idea of importance.”

Simply put, links are love. A link from one site to another sends love. PR is based on the amount of love received.

This diagram shows our interpretation of PR. Here:

  • each sphere is a web page (Page 1,2,3 and 4)
  • sphere size shows relative PR
  • each heart shows love being sent
  • heart size shows relative amount of love

Page 1 and 2 send love to Pages 3 and 4, respectively. Page 3 sends love to Page 4.

Pages 1 and 2 receive no love so:

  • their PR is low
  • the amount of love they send is small

Page 3 receives love from Page 1 so:

  • the PR is higher than Page 1 or 2
  • it sends more love to Page 4

Page 4 receives the most love.

In this system, Pages 1 and 2 would have the lowest PR, Page 3 would be somewhere in the middle and Page 4 would enjoy the highest PR of all.

While the math is complex with many more variables involved, the idea behind PageRank is simple, democratic and beautiful. This technology coupled with a clean interface helped Google dominate the search industry.

PR is useful when evaluating content-based web pages but what about dating sites?

Measuring the Value of a Dating Site

To measure the value of a website, we usually consider product provided and goal achieved. Sites that provide a desired product or help you achieve a goal have value.

On content-based websites:

  • product = online content (exists in cyberspace and can be easily linked to)
  • goal = to give you desired content quickly (occurs online)

On dating sites:

  • product = people (live and breathe in meatspace and cannot be linked to)
  • goal = to get you a date quickly (occurs offline)

With this in mind, GiveYourHandaBreak.com values dating sites in terms of people and your odds of finding a date. Our favorite question is:

What site gives you the best odds for finding a date quickly?

Traffic data can help us answer this question. In particular, the following four metrics:

  • people per month
  • visits per month
  • average length of stay per visit (in minutes)
  • average number of pages viewed each visit

Why are these metrics important?

  • more people per month = greater chance of finding someone in your geographic area
  • more visits per person = greater chance of making contact / less time waiting for reply
  • longer stay per visit = more time spent browsing profiles and communicating
  • more pages viewed per visit = more profiles viewed / greater chance of your profile being viewed

Dating Site Without People = Bookstore Without Books

A site may offer an advanced matching algorithm, robust search engine and free video chat but if it’s only visited by 20 people per month, it is worthless. Why? Statistically speaking, you wouldn’t be able to find a date within 50 miles of your geographic location.

A beautiful dating site without people is like a beautiful bookstore without books.

A site may be visited by 1,000,000 people per month but if each person visits only once per month for less than 10 seconds per visit, it is worthless. It has no value to dating consumers. Why? Statistically speaking, the chances of these casual visitors staying long enough to view your profile would be next to nothing. Even if you emailed them, it would take weeks or months before you received a response because they spend so little time on the site each month.

The value of a dating site is tied to the quantity and quality of traffic it receives.

To see if PR can be used to evaluate dating sites, we need to compare it to traffic data.

PageRank and Traffic Comparison

This table shows the Google PageRank of 122 dating sites. For comparison purposes, we included the dating site category, people per month, visits per month, average stay (in minutes) and pages per visit for each site.

Click table headers to sort. Click dating site name for detailed review and category name for detailed comparison.

Dating Site Category PR People/
Month
Visits/
Month
Stay/
Visit
Pages/
Visit
AfroIntroductions.com Ethnic 4 43,541 219,742 31:08 46.2
AgeMatch.com Niche 4 15,151 39,660 17:06 24.3
AmericanSingles.com General 4 569,092 1,907,353 20:10 89.8
Amigos.com Ethnic 6 226,442 595,725 11:38 17.4
Amor.com Ethnic 4 201,085 552,979 06:17 8.1
AnimalAttraction.com Niche 5 34,980 124,189 04:32 14.5
ArabLounge.com Ethnic 4 44,355 111,406 16:06 24.7
AshleyMadison.com Niche 5 501,982 1,809,581 11:48 29.3
AsiaFriendFinder.com Ethnic 6 116,268 423,088 14:09 18.0
AsianPeopleMeet.com Ethnic 4 89,017 409,961 09:57 46.8
BabyBoomerPeopleMeet.com Niche 3 108,448 695,046 07:48 29.1
BBPeopleMeet.com Niche 3 227,555 1,616,587 12:11 42.4
BharatMatrimony.com Ethnic 6 64,962 100,636 02:00 2.8
BigChurch.com Niche 6 96,852 330,863 14:33 25.3
BikerKiss.com Niche 4 51,718 159,654 10:40 17.5
BlackBabyBoomerMeet.com Niche 3 42,716 273,458 12:59 38.9
BlackChristianPeopleMeet.com Niche 3 98,654 570,826 13:52 43.9
BlackPeopleMeet.com Ethnic 5 617,218 4,658,999 21:00 65.1
BlackSingles.com Ethnic 5 415,149 1,462,968 14:13 23.7
Bookofmatches.com Free 4 182,346 511,823 15:44 41.8
CasualKiss.com Free 3 12,686 97,690 41:14 252.0
Chemistry.com General 7 744,448 2,466,748 10:09 22.7
ChineseFriendFinder.com Ethnic 5 2,870 3,787 09:37 8.5
ChristianCafe.com Niche 4 165,967 730,270 14:59 23.4
ConnectingSingles.com Free 4 75,242 438,570 08:49 17.1
CrazyBlindDate.com Free 5 15,413 43,305 06:37 15.4
Date.com General 5 380,751 1,412,378 09:53 19.5
DateAMillionaire.com Wealthy 4 51,580 120,658 10:41 29.0
DateHookup.com Free 5 804,892 7,475,542 12:25 21.8
DateInAsia.com Ethnic 4 18,933 297,775 23:17 39.7
DateMyPet.com Niche 4 6,687 13,975 25:43 40.5
Deafs.com Niche 4 10,087 23,854 15:06 35.9
DemocraticPeopleMeet.com Niche 3 41,880 105,911 04:35 16.9
DivorcedPeopleMeet.com Niche 2 54,532 247,325 05:39 25.0
EbonyFriends.com Ethnic 4 2,679 3,982 05:52 8.7
eHarmony.com General 7 4,167,975 20,582,651 11:37 17.2
Engage.com General 7 49,868 67,700 05:46 9.0
EquestrianCupid.com Niche 4 11,843 27,223 13:13 15.0
EquestrianSingles.com Niche 4 36,644 267,120 16:28 53.1
EWSingles.com Niche 5 2,216 4,498 07:23 13.2
FarmersOnly.com Niche 4 42,859 211,275 13:47 17.0
FastCupid.com Niche 4 39,672 121,464 03:38 5.2
FilipinaHeart.com Ethnic 4 60,826 481,146 27:14 44.0
FilipinoFriendFinder.com Ethnic 5 90,236 211,619 12:28 21.9
Fitness-Singles.com Niche 4 209,701 715,663 11:51 30.4
FrenchFriendFinder.com Ethnic 6 9,015 10,562 03:18 4.4
FriendFinder.com General 6 522,927 1,251,795 09:07 19.7
GermanFriendFinder.com Ethnic 5 19,609 33,477 05:43 6.2
GothicMatch.com Niche 4 12,201 23,784 07:10 7.9
GreenFriends.com Niche 5 2,644 3,730 01:35 2.8
GreenSingles.com Niche 5 25,484 130,435 07:49 14.2
HorseandCountrySingles.com Niche 3 9,546 20,451 08:38 18.1
HOTorNOT.com General 6 849,873 3,732,027 09:53 35.1
IndiaMatch.com Ethnic 3 39,681 124,245 08:57 28.8
IndianFriendFinder.com Ethnic 5 36,220 61,242 08:57 44.0
IntellectConnect.com Free 3 1,692 1,692 07:39 22.5
InterracialMatch.com Ethnic 4 26,134 56,377 11:46 17.1
InterracialPeopleMeet.com Ethnic 3 38,646 273,730 12:45 38.9
IranianPersonals.com Ethnic 4 31,579 138,391 22:22 35.2
ItalianFriendFinder.com Ethnic 6 14,160 38,727 02:37 3.7
JDate.com Ethnic 6 184,754 1,454,440 32:13 132.6
JewishCafe.com Ethnic 5 8,125 23,571 11:49 17.9
JewishFriendFinder.com Ethnic 5 13,256 19,699 04:59 5.9
JPeopleMeet.com Ethnic 3 22,376 208,533 13:39 24.9
KoreanFriendFinder.com Ethnic 5 24,220 51,149 05:05 7.0
LargeFriends.com Niche 4 35,325 101,883 13:09 16.7
LatinAmericanCupid.com Ethnic 4 103,371 297,174 21:53 24.7
LatinoPeopleMeet.com Ethnic 4 81,655 430,385 12:54 41.6
Lavalife.com General 5 423,282 1,592,193 11:15 20.0
Lets101.com Free 4 12,906 24,001 09:46 15.8
LoveAndSeek.com Niche 3 254,288 1,662,588 13:41 45.1
Lovebyrd.com Niche 4 3,685 9,078 25:01 22.4
LoveFromIndia.com Ethnic 4 253,973 318,533 02:24 12.0
Match.com General 7 4,938,370 28,337,925 13:55 22.2
Matchdoctor.com General 5 168,905 1,206,263 02:38 18.5
Matchmaker.com General 7 439,806 1,230,690 07:46 11.5
Mate1.com General 4 1,469,840 5,018,213 03:19 9.2
Meeta.com Free 4 5,940 12,425 03:03 4.9
MeetMarketAdventures.com Niche 5 11,637 16,986 06:07 5.8
Metrodate.com General 5 45,789 77,611 11:01 9.7
MilitaryFriends.com Niche 4 6,617 12,145 07:00 12.3
MillionaireMatch.com Wealthy 5 65,240 308,146 12:48 19.3
MillionaireMate.com Wealthy 5 81,310 155,342 09:00 8.4
Mingle2.com Free 6 280,280 2,132,577 11:14 18.8
Mingles.com Free 3 43,828 133,153 14:27 28.2
MuslimFriends.com Ethnic 3 4,861 6,832 10:58 22.2
MuslimMatrimonial.com Ethnic 5 4,263 16,810 08:05 8.6
MuslimMatrimonialsNetwork.com Ethnic 3 8,718 55,083 13:07 25.9
Naseeb.com Ethnic 5 4,291 7,010 08:33 11.0
OkCupid.com Free 6 632,660 4,293,936 17:17 31.1
PassionsNetwork.com Free 4 16,170 28,476 06:08 8.1
Perfectmatch.com General 6 483,174 1,096,479 08:51 15.2
PetPeopleMeet.com Niche 3 22,242 107,981 07:33 24.5
PetsFriends.com Niche 3 2,893 2,893 05:58 4.6
Plentyoffish.com Free 6 2,962,299 36,179,852 18:55 32.0
Qiran.com Ethnic 4 17,482 66,735 14:46 25.8
RepublicanPeopleMeet.com Niche 3 24,454 91,063 05:49 21.9
SeekingArrangement.com Wealthy 3 43,418 152,799 12:51 27.5
SeekingMillionaire.com Wealthy 3 18,736 59,148 12:03 19.7
SeniorBlackPeopleMeet.com Niche 2 15,068 157,618 07:30 26.1
SeniorFriendFinder.com Niche 5 377,881 1,282,023 07:03 15.6
SeniorMatch.com Niche 4 10,768 26,014 08:59 13.5
SeniorPeopleMeet.com Niche 3 302,881 2,159,647 10:25 37.3
Shaadi.com Ethnic 6 337,260 794,823 07:51 15.3
SingleParentMatch.com Niche 4 3,821 4,374 17:13 17.9
SingleParentMeet.com Niche 3 366,861 1,950,549 12:28 44.0
SinglesCrowd.com General 4 25,702 32,549 04:40 5.6
Singlesnet.com General 4 8,498,684 42,767,403 10:48 25.4
Smooch.com Free 3 21,824 198,534 07:55 13.5
SoulSingles.com Ethnic 3 35,545 95,902 35:01 100.1
SpeedDate.com Niche 5 217,377 514,954 02:53 9.8
Sugardaddie.com Wealthy 5 108,575 537,071 11:28 26.1
SugarDaddyForMe.com Wealthy 4 163,295 613,252 12:48 19.9
TallFriends.com Niche 4 4,727 15,601 06:54 13.5
TeamDating.com Free 4 1,463 1,463 02:00 1
TheAtlasphere.com Niche 5 5,965 10,575 03:32 2.9
True.com General 5 3,410,791 12,793,444 08:14 11.2
uDate.com General 7 37,868 82,342 12:37 23.8
VeggieConnection.com Niche 4 3,336 8,727 07:27 12.4
VietnamCupid.com Ethnic 4 29,437 85,452 12:34 15.0
WealthyMen.com Wealthy 5 2,647,693 5,582,673 00:37 1.6
Webdate.com Free 5 759,274 1,935,659 08:44 14.3

To keep things as objective as possible, we relied on traffic data from a third-party provider.

Compete triangulates multiple data sources, including ISP, Panel & Toolbar to estimate U.S. traffic. It is a free service that provides traffic information, including unique visitors per month, visits, average time spent and pages viewed per visit. The data were collected between the middle of last month and the beginning of this month so they reflect the traffic during the month of September, 2008.

PageRank data were collected using the Google Toolbar between September and November, 2008. Please note that the PR shown is the front-page PR. Also note that PR changes constantly and may have changed since we last checked.

Observations: Category

Natural Link Love?

If you sort our dynamic table by PR (from highest to lowest), the first thing you will notice is that sites in the “general” category dominate the top of the PR ladder. Of the 20 sites with PR 6 and 7, there are:

  • 9 general (45%)
  • 7 ethnic (35%)
  • 3 free (15%)
  • 1 niche (5%)

If you consider the total number of sites in each category, however, the order changes. What percent of sites in each category made it into the top 20 PR?

  • 9/18 general (50%)
  • 3/16 free (19%)
  • 7/38 ethnic (18%)
  • 1/42 niche (2%)

General sites have incoming links from a much broader spectrum of sites so they would have a higher PR than niche sites.

Free sites may get a PR advantage from having more “natural” incoming links. Non-free (paid subscription) sites are generally more profitable and can afford to advertise much more aggressively. While this helps bring traffic to their site, these “artificial” (affiliate or paid advertising) links do not help their PR. The “natural” link love that free sites seem to attract, on the other hand, would boost PR.

Why were there so many ethnic sites in the top 20?

Well, 6 of the 7 ethnic sites are part of the FriendFinder Network. This network has been around since 1996. Older sites tend to have more incoming links and thus would enjoy a higher PR. Cross-linking between sites on the FriendFinder network (from older sites to newer sites) would help, too.

Observations: People/Month

At first glance, we saw no correlation between PR and people per month:

  • there were PR 7 sites with as little as 40,000 people/month
  • the site with the most people per month (over 8 million) had a modest PR of 4

On closer inspection, we noticed a pattern.

The following table shows the minimum, maximum, mean and median of people/month for each PageRank. For reference, we included the number of sites.

PR Sites Minimum Maximum Mean Median
PR 7 6 37,868 4,938,370 1,729,723 592,127
PR 6 14 9,015 2,962,299 484,352 253,361
PR 5 32 2,216 3,410,791 353,986 55,515
PR 4 44 1,463 8,498,684 289,708 33,452
PR 3 24 1,692 366,861 74,768 37,096
PR 2 2 15,068 54,532 34,800 34,800

By examining the mean and median, we saw that sites with more people/month tended to have higher PR.

Why? Does Google consider unique visitors when calculating PR or do sites with more traffic simply gather more incoming links? We don’t know. Like any good recipe, Google keeps the ingredients of the PageRank formula a trade secret.

All we know is this:

  • in general, sites with more people/month had higher PR
  • high PR does not necessarily mean more people/month
  • more people/month does not guarantee high PR

Observations: Visits/Month

Again, first impressions were deceiving:

  • there were PR 7 sites with as little as 67,000 visits/month
  • the site with the most visits (over 42 million) had a PR of 4

Closer inspection revealed the pattern.

The following table shows the minimum, maximum, mean and median of visits/month for each PageRank.

PR Sites Minimum Maximum Mean Median
PR 7 6 67,700 28,337,925 8,794,676 1,848,719
PR 6 14 10,562 36,179,852 3,745,395 945,651
PR 5 32 3730 12,793,444 1,361,467 142,889
PR 4 44 1,463 42,767,403 1,305,863 116,032
PR 3 24 1,692 2,159,647 444,348 128,699
PR 2 2 157,618 247,325 202,472 202,472

Again, the mean and median told us that sites with more visits tended to have higher PR.

Why? Maybe Google considers visits when calculating PR or maybe sites visited more often attract more incoming links. Again, we don’t know.

What did we learn?

  • in general, sites with more visits/month had higher PR
  • high PR does not necessarily mean more visits
  • more visits/month does not guarantee high PR

Observations: Stay/Visit

A quick glance led us to suspect no correlation between stay per visit and PR:

  • there were PR 7 sites with an average stay/visit of less than 6 minutes
  • two sites with excellent average stay (35 and 41 minutes) had a low PR of 3

Examining the mean and median data confirmed our suspicion.

The following table shows the minimum, maximum, mean and median of average stay per visit (in minutes) for each PageRank. For ease of calculation, we rounded off to the nearest minute.

PR Sites Minimum Maximum Mean Median
PR 7 6 6:00 14:00 10:30 11:00
PR 6 14 2:00 32:00 11:42 10:30
PR 5 32 1:00 21:00 8:24 8:30
PR 4 44 2:00 31:00 12:48 12:30
PR 3 24 5:00 41:00 12:48 12:00
PR 2 2 6:00 8:00 7:00 7:00

We saw no correlation between PR and average stay per visit.

Observations: Pages/Visit

A quick glance suggested no correlation:

  • there were PR 7 sites with an average pages viewed per visit of only 9 pages
  • sites with an average of over 100 pages/visit had a PR of 3

The data supported this.

The following table shows the minimum, maximum, mean and median of pages viewed per visit for each PageRank.

PR Sites Minimum Maximum Mean Median
PR 7 6 9.0 23.8 17.7 19.7
PR 6 14 2.8 132.6 26.5 18.4
PR 5 32 1.6 65.1 16.1 14.3
PR 4 44 1.0 89.8 23.6 17.7
PR 3 24 4.6 252.0 40.5 27.9
PR 2 2 25.0 26.1 25.6 25.6

We saw no correlation between PR and average number of pages viewed per visit.

Charts

To make it easier to see the relationship between PR and each metric, we exported the data from the above tables as bar charts (shown below).

(Mouse over to see our comments…)

Hopefully this visualization will make things clearer. We thank Mr. Santos for his kind suggestions!

Conclusion

To sum up our observations:

  • general dating sites tended to have higher PR than niche sites
  • older sites seemed to have a PR advantage (even for low traffic sites)
  • sites that are part of a large network seemed to enjoy higher PR (even for low traffic sites)
  • sites with more people/month tended to have higher PR (but there were exceptions)
  • sites with more visits/month tended to have higher PR (but there were exceptions)
  • we saw no correlation between PR and average stay/visit
  • we saw no correlation between PR and average pages viewed/visit

As we said earlier:

The value of a dating site is tied to the quantity and quality of traffic it receives.

While there seems to be a correlation between PR and quantity of traffic (people/month and visits), we found no correlation between PR and quality of traffic (average time spent and pages viewed per visit).

Average stay is directly proportional to time spent communicating and pages viewed is directly proportional to profiles viewed.

Both are essential traffic data for measuring the value of a dating site.

Therefore, we conclude that:

Based on the above data, PageRank alone is not enough for evaluating dating sites.

In addition to PR, it is essential that you check our Dating Site Web Statistics page to give you an idea of traffic quality.

Additional Findings

Recently, we found another way PR can be very useful with regards to online personals (or any industry for that matter).

Last month, we removed a dating site because:

  • the page title and URL no longer matched
  • most links were pointing to a second (external) site
  • both sites looked like identical content on different domains (!)

What brought our attention to these inconsistencies?

The site’s PageRank dropped from PR 5 to PR 0 within a short time (several months).

That leads us to our second conclusion:

Based on past experience, changes in PR can be useful as a warning signal.

Sudden drops in dating site PR may indicate deeper problems that warrant further investigation.

Related Articles:
Unique Visitors Metric: Not Enough on Dating Sites? (http://www.giveyourhandabreak.com/articles/unique-visitors/)
Millions of Members: Medal of Honor or Badge of Shame? (http://www.giveyourhandabreak.com/articles/millions-of-members/)

Related Pages:
Dating Site Web Statistics (http://www.giveyourhandabreak.com/web-stats/)

Sources:
Web Stats by Compete Site Analytics (http://siteanalytics.compete.com/)
Google PageRank by Google Toolbar (http://toolbar.google.com/)

Internal Resources:
Articles on Dating Sites & Online Dating (http://www.giveyourhandabreak.com/category/articles/)

External Resources:
Google (http://www.google.com/)
Google PageRank Technology Explained (http://www.google.com/corporate/tech.html)
Original Google Paper (http://infolab.stanford.edu/~backrub/google.html)
Compete Metric Descriptions (http://www.compete.com/help/s2)

Comments are owned by whoever posted them. We are not responsible for them in any way.

Post a Comment