Squidoo LensRank Factors: A Study

The Squidoo FAQ says:

We use an automated algorithm-- LensRank-- to rank the lenses. We look at community ratings, lensmaster reputation, clickthrough rates, frequency of updates, inbound and outbound links, revenue generated, and lots of other factors and give the lens a number.

This part of the FAQ hasn't changed in years, so it may not be current. But to break it down:

• Community ratings include SquidLikes and Angel Blessings. Community ratings used to include Favorites and perhaps Lensroll, but those factors were retired in Jun '11.
  


• Lensmaster reputation used to be based on a lensmaster's fan club (like "followers" on Twitter). Fan clubs were retired this spring along with Favorites. It's possible that Gianthood is counted as "lensmaster reputation," but I'm doubtful. Nor do I think Monster levels are factors: they don't reflect a lensmaster's ability to write good lenses that people like. Trophies such as "number of comments received" might count as reputation, or perhaps this lensrank factor has been phased out.


• I assume Clickthrough Rates means the ratio of visitors to clickouts. In other words, how many people found something on your page worth clicking? Fluff's SquidUtils Workshop Add-On adds CTR to our Lens Stats "Traffic" pane.


• Frequency of Updates sounds like it's measuring how often a lens is updated. I don't edit/update lenses on a regular enough schedule to be sure. All I know is that starting this last spring, the lensrank jump has been greater when I republish a lens, and lensrank decays faster. (I hate this factor... quality content is quality content, grr!).


• Inbound links means: Which links on other webpages bring traffic to the lens?


• Outbound links means: Which links do people click to leave the lens?


• Revenue generated: We usually assume this means revenue from Amazon and other affiliate sales. It might also refer to how much ad income is generated. Squidoo doesn't reward clicks on adds directly, as that would lead to clickfraud, but perhaps all revenue-earning clicks get a small lensrank boost.


• Other factors include a one-time lensrank boost from purple stars, and Lensrank Turboboosts unlocked as prizes at a few levels in the Monster Points game (See Fluff's useful post: Which Lens Should Get the TurboBoost?). Years ago, Giants who were accepted to the old Giant 100 Club could select one lens to receive a Gold Star and a 20% lensrank boost. I don't see this perk listed on the updated Giant FAQ, so it may have been retired. It may be grandfathered (see below).

So that's the official word on Squidoo lensrank. Everything from here on out is my speculation. Now let's take a look at my dashboard. Or, if you want to skip the in-depth discussion, see my summary of OBSERVED lensrank factors at the end.

Here's the stats of ALL my tier ones, ALL my tier twos, and some of my top tier threes.

I once assumed that all stats on the default Squidoo dashboard -- Likes, Weekly Traffic, Earnings, Last Update -- mattered as lensrank factors.

However, I no longer think that's so. I've seen the impact of Likes (ratings, thumbs by Squidoo members) wear off after a week or so...maybe ten days? It's hard to say. As a result, I have lenses with 100+ Squidlikes in the lensrank 125,000 range.

So I think recent likes matter for lensrank, but total likes don't matter. Except, perhaps, in this way: ALL of my tier one and tier two lenses have several Squidlikes, and these days that's true for nearly all of tier three. I don't know if that's cause (no Squidlikes means a zero in one part of the lensrank formula, which is bad) or effect (a successful lens tends to get likes).

The other factors on the main pane of our dashboard stats do matter for lensrank, although we don't know by how much, or how quickly they wear off. (Note that sales, likes and blessings seem to kick in a lensrank boost only after a 2-3 day lag.)

Clickouts are a lensrank factor that's not on our default dashboard. To see clickouts, click "Stats" under a lensname. That "stats summary" pane tells us 30-day clickouts.

I've noted in bold what look like anomalies: lenses with more or less traffic than I'd expect for their lensrank.

My top lens may still be enjoying its old Gold Star boost. On the other hand, it receives a steady stream of likes, thanks to being featured on my profile and as my top lens listed in my bio box on my 235+ lenses. Other factors may be in play. (Rumored, but never confirmed: "Time spent on lens": visitors tend to read that lens all the way through, and it's a long one.)

Also notice lens "YC," tenth row: nearly 1300 clickouts, 150-200 visits a week, but even an exceptional number of clickouts cannot keep that lens in tier one.

Pages/visit means how often visitors come back to a lens. If this is one of those "other" lensrank factors, it may partially explain lens #1's great lensrank; it has the most return visits of any of my lenses. Nearly all my top tier lenses have Pages/visit near or above 1.50+.

"Outlinks" is my shorthand for "outbound links," specifically the number of DIFFERENT outbound links that visitors clicked on. This stat and "CTR" (Click-through-rate) are supplied by SquidUtils' workshop add-on, NOT by Squidoo. I'm monitoring them because Squidoo's FAQ mentions "outbound links" and "clickthrough rates" as lensrank factors.

Does the FAQ's "outbound links" mean (a) total number of outbound links on a lens, regardless of whether anyone clicks them, (b) total number of clickouts or (c) how many different outbound links visitors clicked? I'm going with (c), because counting (a) encourages spammy lenses, while the FAQ mentions clickouts (b) separately. If people are clicking on lots of different links on a lens, that's a good sign they're finding something worthwhile.

Unfortunately, I'm not seeing any major patterns jumping out with CTR and outbound links. Maybe the Squidoo FAQ is out of date (like THAT never happens), or maybe they're just not as important as other factors. But now I'm wondering if Pages/visit -- visitors leaving and returning in the same session -- may help lensrank.

Squidoo monitors community ratings and visitor participation, in order to detect lenses which may have good information and content, but which may not happen to "tick off" factors that a dumb computer algorithm can measure.

First, let's look at community ratings: likes and blessings.

Squidoo-related lenses tend to get more Squidlikes. Notice one of my traffic tips lenses (SS) down at position 13. It always has lots of comments and likes, but it's fallen out of tier one as competition has grown. That fits my impression that many Squidlikes can get a lens into tier two, but they can't boost it to tier one without other factors. (I once had a lens go viral and earn 68 Squidlikes in a week, but it still didn't crack tier one.)

Similarly, I think isolated blessings aren't enough to move a lens up into a tier unless it's on the cusp already, but they may keep lenses from slipping. Often, the lensrank boost from a single blessing is so minor I don't see it: other "ranking signals" seem to drown it out. However, whenever there's an Angel quest, I see a noticeable lensrank drop two days later, as if hundreds of lenses being blessed has caused significant lensrank churn. (Many of those lenses may also be getting Angel comments, likes, clickouts and other interaction.)

Not reflected on this chart is the fact that lensrank boosts from likes and blessings wear off quickly: I haven't tracked it precisely, but my impression is a week to two weeks.

So, what about visitor participation?

I was pleasantly surprised to see that Poll-taking and quizzes have little impact on lensrank. I have a batch of popular quizzes, but only two of them showed up on this chart -- that's TS and TQ -- and both are tier 3. The rest are lower. Likewise, one lens with a bunch of polls (XF, #44) is nowhere near tier two. It's probably just as well polls and quizzes don't count much for lensrank: SquidPoints hunters create and take quizzes and polls to rack up monster points, and the points hunting game doesn't have much to do with lens quality.

I don't think Comments are a major factor, either. Notice #13 again: lots of comments and likes in the last week, but without mega-traffic, clickouts or sales, it's stuck in tier two.

The Squidoo FAQ mentions "inbound links" as a lensrank factor, but how does Squidoo measure them? Does it use Yahoo Site Explorer, as Squidaholic does, to count backlinks?

Surely not. Every SEO tool finds a slightly different set of different backlinks, so the count is somewhat arbitrary. Nor do search engines count all links as equally important: a forum signature is a self-advertisement repeated on hundreds of posts, whereas a link from CNET is an unsolicited recommendation. Finally, not all backlinks are relevant to a particular search: a backlink from the Old Spice guy's YouTube Channel is relevant to viral videos, not nuclear physics. Counting backlinks, in short, won't tell you which links matter to search engines.

On the other hand, counting links that have actually attracted visitors is a measure of lens performance, and Squidoo's detailed traffic stats measure these "inbound links" in a few ways. The Traffic pane of lens stats reports how much total traffic is coming from links ("referrals") AND how many different links are sending traffic ("referrers"). I've tallied up the number of different "referrer" domains on my chart, trying to see if Squidoo treats diversity of referrers as a lensrank factor. It may, but if so, it's not much.

I also checked the number of different search engines (SEs) sending traffic, and saw no impact on lensrank whatsoever. Likewise, I checked to see if a variety of different searches (keywords) might be a lensrank factor. I don't think so: I see lenses getting traffic with lots of different search phrases that are down at the bottom of the chart.

Finally, I checked to see if lensrank counted direct traffic as significant. This is traffic from someone entering a link manually, using a bookmark, or clicking a link in email. I thought perhaps Squidoo would consider that a vote of confidence. I see lenses with high direct traffic tend to be tier one and two lenses, but I'm not sure whether this is cause (a minor lensrank factor) or effect.

Here's some lensrank factors that the dashboard seems to treat as significant for a longer span of time. Does lensrank think clickouts are important for 30 days, or does the dashboard just pick that span by coincidence?

The other question is how long sales matter. I don't think their lensrank boost lasts the full two months after a sale; the pending "$$$" column is there to tell us of possible commissions. From what I've seen, it's more of a week-to-ten-days type boost. Or maybe two weeks.

I mostly wanted to show you this chart to draw attention to lens HC down in row 37. It's had 14 sales in the last month, yet it's tier three. This lens shows the fact that successful sales lenses often attract a limited, targeted audience rather than lots of general-interest traffic. Lots of people may visit a page with cute kitten pictures and help it rank well, but you're more likely to get a sale if your lens only attracts visitors interested in buying the organic kitten food you're reviewing.

Tweets, Facebook likes, Google+ may impact lensrank indirectly by bringing a burst of traffic from your followers. Also, search engines are beginning to experiment with letting social media factors like this influence their search engine rankings a tiny bit. But does Squidoo measure Tweets, FB likes and so on as lensrank factors?

I don't think so for a few reasons. First, as I noted above, Squidoo's FAQ says lensrank measures community ratings, whereas most FB likes and Tweets come from outside the Squidoo community. It would be too easy to game the system, if one had a ton of friends on a social network, asking everyone to come rate your lens. Traffic exchanges and lensrank swaps are against Squidoo's TOS, and it's a lot easier to catch them on-site than off-site. So, I see problems in counting off-site ratings for on-site lensrank.

Also, Squidoo is always fiddling with which social media buttons appear at the top of lenses. One week we'll have Facebook up there. Another week it'll be Digg, StumbleUpon, or +1. I haven't noticed a big lensrank upheaval every time those buttons are rearranged, added or removed, so I don't think they're part of the equation.

There's one last factor to note: creation date. My chart above shows that old, established lenses can still compete with new lenses, and new lenses can race to the top tier, but in either case it's very, very difficult for them to stay there.

So that's all the data I've got. Let me summarize my conclusions (such as they are) below. But first, for Excel users, here's the whole chart NOT broken up into chunks, so you can compare all these lensrank factors side-by-side. Maybe you'll see something I missed:

Traffic: Officially confirmed.
Clickouts: Officially confirmed.
Recently updated/republished: Officially confirmed.
Sales: Officially confirmed.
Amount of sale: Doesn't appear to be a factor.
RECENT Squidlikes: Officially confirmed.
Lifetime Squidlikes: Probably not a factor.
Blessings: Officially confirmed.
Purple Stars: Officially confirmed, but the impact seems very minor.
Lensrank turboboost: Officially confirmed, but negligible.
Clickthrough rates (CTR): Officially confirmed, but appears minor.
Variety of outbound links: Perhaps a minor factor.
Gold Stars (grandfathered): Likely.
Pages/visit: No official info on this, but circumstantial evidence.
Comments: Maybe, but if so, minor factor.
Polls and quizzes: Maybe, but if so, minor factor.
Frequency of updates: FAQ says so; I haven't seen it. (Frequency isn't the same as "was it republished in the last few days?" which is a factor.)
Time spent on lens: Unconfirmed, but there's hints this is a factor.
Social Media (Tweets, Facebook, +1): Little or no direct lensrank correlation observed, although of course Tweets and FB bring visitors.
Clicks on ads: Unconfirmed, but possible.
Variety of traffic sources: My hunch, but no strong evidence for this.
Variety of searches/kewords: No.
Creation date: Not a factor.

A lens must excel in many of these factors to reach the top two tiers, and it has to keep getting new sales, likes, clickouts, or other "boosts" to stay up there. Consistently high traffic isn't enough.

Icons from mysitemyway.com.