Correlation Data for SEO and Social Media Analysis – Part 1 – Whiteboard Friday


howdy SEOmoz supporters therefore welcomed another publication of whiteboard Friday today we’ve got a great topic too exciting topic we’re gonna be talking about equivalence data and how you can use it in SEO and social media analysis so a lot of you already use correlation data and many of you have probably seen on our blog and in our early freeing of data from our rank influences questionnaire this year that we’re going to be presenting a assortment of correlation data we’ve done this a number of times over the past few years it’s always particularly very interesting to look at sometimes it’s potentially contentious because some of the data is really interesting and surprising and in this case correlation data is something that I know numerous kinfolks in the SEO field in the social media orbit who don’t have you know a substantive background in statistics like me I imply let’s face it I repute I got a D in my statistics class in college and I’m pretty sure I bounced the last three castes and didn’t even go to the final because I was sure I was gonna neglect and somehow I skated by but that’s beside the point I drooped out of college anyway so I didn’t I didn’t need that D but correlation it it sounds like a big fancy word but it’s actually really really simple it’s essentially the extent to which one metric prophesies or has a correlation a connection to another so let me give you a couple of really simple lessons just so you can understand this and then we’ll talk about some ways to use it in SEO and social so let’s imagine for a second that you are a contractor right and you’re doing some some material writing and you improved those enterprises that you write for on an hourly basis so you do 10 hours of strive right at 10 dollars an hour or let’s say and you get a hundred dollars so the hours build and the dollars received have a extremely very good correlation in fact they’d have a correlation of 1.0 hopefully right hopefully you don’t construct some hours and then beings don’t remunerate or they compensate you more hours than you construct perhaps perhaps those things will happen but often it’s a one-to-one correlation its 1.0 as the linkage because retain all linkage multitudes at least statistically speaking it from a math attitude are between 0& 1 positive equivalences at least right and then we do have negative linkages as well we’re not going to worry about those for a sec but in the in the dollars received hours improve that’s a as a perfect linkage right so you’d see I improve for 1 hour I came $10 I build for two hours I went $20 I construct for 3 hour right so perfect linear nice one point zero correlation you can imagine there are there are lots of systems simple organisations that perform like this for example number of steps that you take and the interval that you circulate right the issue is those so have kind of a perfect delightful linkage and then there’s trash that’s has a correlation but the linkage might not be as perfectly predictive and we want to have numbers around what those linkages are so here’s a pretty simple example right this is the number of daytimes that I wear yellowish shoes you can see I’m not wearing them today but number of eras that I wear yellow shoes and the number of daytimes where I yield a professional presentation and oftentimes these are quite connected right but it turns out there’s also days where I wear the yellowish shoes and I don’t yield a exhibition or where I impart a present but I don’t wear the yellow shoes those those things do happen right it’s not a requirement that every time I I get up on stage after wear yellow shoes but it happens a lot right and so we can map those we can say oh well here’s you know there was five days where ran wore yellowed shoes and and all five of those daytimes he gave a presentation and then a duet days later oh you know what feed wore the yellow shoes merely around township right he was breaking in a brand-new duet and so there’s a marry more eras where he wore them but only one more day where you leave a show so we get a little you know chart like this and what linkage compositions can do is they can help present a number like 0.72 the ties between these two quantities so you can sort of say huh well there’s a good correlation between them but it’s not certain that every time rolled wears yellow shoes he’s causing a lecture or each time he passes a demonstration he wears yellow shoes and that’s exactly what these multitudes are designed to predict now in really simple situations like this right a linkage compose of 0.7 oh it’s relatively high claim but we’d actually need quite a few data points to be able to predict something announced standard error so standard error tells us the degree to which we’re certain that these two things are connected right so if we have a standard error of let’s say you know 0.25 that might be a quite high standard error because we only have a few data points means that there’s potentially a great deal of wavering this could be a much lower correlation than we think it is or far higher one depending but if we got thousands of data points if we had every data point around when I worn yellowish shoes every data point around what I’ve given a professional presentation this standard error might sag dramatically to let say level zero five right and now we can be more certain that oh yeah there’s there’s clearly a connect there and with a little bit of fluctuation we know pretty much what the equivalence digit is so we can predict how often when torrent returns a demonstration he’s gonna be wearing amber shoes based on an average of previous data right and so that’s what this is designed to tell us that’s exactly what connect can be used for so let’s talk about some ways to use correlation data in your SEO and your social media campaigns first off in in a lot of the cases you don’t actually need a huge data set let’s talk first about behaviors that you’re probably previously utilizing correlation data which is with individual data points right these are things where you gather you look at search results or you look at how you being carried out in social media you look at how other people are doing and you pattern correlations in your mind like son you know what every time I examine someone write a top ten list about something that seems to get a lot of joins and a great deal of retweets and a lot of attention it seems like top X registers are a really good way to produce content that parties really like these top top ten schedules or surface X directories maybe maybe that’s a good way to go and that type of data point connection in your own mind is linkage right it’s something where you’re link these things seem to predict success and so I am going to potentially imitate them and see if they prophesy success for me and that’s actually a punishment thing to do so you could do something like hmm it seems like when I have a tweet with a tie that gets higher click-through rate it also comes more retweets so if I can figure out the formula to get one of these probabilities are I’ll do well with both of them right and so I’m gonna work you know my click-through rate I’m gonna work on things that prophesy higher click-through rate I’m going to get those short-lived punchy entitlements I’m gonna get a good URL shortener I’m gonna keep the you know what whatever it is that the format of the tweet that you move that gets one of these is you can you can generally prophesy you’ll get the other one in something maybe in some cases right so this doesn’t undoubtedly apply to everyone a lot of the time it’s just your personal experience and that’s a punishment thing to use or Facebook shares right so you might notice that in your Facebook account when you share content that has a picture of a human face right so it’s got a little oh look there’s a nice picture of Rand I I materialize relatively stick-figure II today yes I glean like a second grader it’s it’s weird that I do whiteboard Friday right but so Facebook shares that have a human face as the thumbnail come more clinks and so you think you’re so ha alright maybe I need to start using more human faces in the thumbnail of what I put on Facebook you know the image that you choose when you share content on there and that might be a fine thing to discover you could use that from an suspicion basis or you could actually measure it right you could go back through your detail and look at all the click-through rates that you’ve earned if you’re using a URL track or a shortener like bitly and then you could see is this really the example you know articulated the numbers into Excel and race the data see on average how you’re performing pretty simple way to do things you might also notice something like an observational notice so links with keywords in the fasten textbook furnish more of a ranks improve in Google right so when you get tie-ups external links and they contain the keyword you’re targeting somewhere in the linchpin textbook that you get more ranking boosters so you think your huh anchor text that must be a powerful signal I’m gonna start trying to do that when I get anchor text on other websites maybe I’m gonna introduced it in my you know in my bio so when people link to me though they’ll call that that special keyword importing the sheets that I want and this observational equivalence is something that SEO and social media marketers and marketers digital purveyors of all stripes have used for ages right they’ve used forever this observational type of correlation but there’s cool stuff that you can do on a research basis that we call you know sort of aggregated or median correlations that produces lots of actually in interesting substance extremely I’ll give you some sample of those so over at HubSpot their social media scientists and Zarrella grows something called the science of retweets talking about how retweets are spread over the web and what associates well with things to become more and more retweets versus less retweets he also does one that’s great on the social sciences of timing talking about when is the best time to tweet or make a blog positions and this equivalence type of data is used all over the place right in tons and tons of different fields clearly in digital marketing and we do some cool trash now at SEO Moz where we collect you know hundreds or thousands of data points right to be able to show aggregate or median correlation with two different metrics so for example right we we might collect 10,000 in our recent survey we collected 10,000 different search results the above reasons we collect such a high one remember is because we want that low-toned low-pitched standard error that comes from a sizzling having a lot of data right so we collect 10,000 and then we insure oh you know how do how do tweets correlate with higher or lower rankings in Google how to face book shares correlate with higher or lower rankings in Google and you can see actually that some of the interesting things we’ve noticed from accumulating this type of data is that hmm key words in the alt attribute of an idol for example predict higher median connect than exercising the keyword in the h1 so a lot of 8th SEO is tell you oh you know that h1 tag that’s really important I got to get the keywords in the h1 tag got to have h1 s on every sheet seems to us like the equivalence with H ones key words in the h1 is no better than having the key word exactly near at the particularly top of the sheet which h1 is usually predict anyway so maybe it’s not the h1 that’ s helping no no privilege it’s connect data it’s not causation you know we don’t know for sure that this is what’s causing it but we know that there’s a connection numerically between these these metrics but that alt property huh it appears seems positive and we’ve never judge oh maybe we should recommend that so for the last few years we’ve been recommending framed a good portrait on the page and make sure your keywords in there you can see we did this with Twitter data we did a cool study with Twitter data where we looked at a large number of tweets and we said what foresees higher click-through rate and turned out that shorter tweets developed higher click-through rate probably no surprise and so instead of using all 140 attributes you exclusively use you are well aware 60 people 80 reputations looks a lot like more beings click on the links in those shorter tweets and that’s kind of interesting kind of cool maybe it suggests that when we’re writing titles and headlines of things we want people to click should obligate those tweets very short we looked at you know putting the the link in the tweet at the figurehead of the tweet versus the end of the tweet versus the centre the middle-of-the-road glances slightly higher than the other two so you can learn all sorts of interesting stuff and this is what’s awesome about correlation data it doesn’t necessarily mean it predicts things but what it does aim is that things that have these features right have a higher propensity to do well so in some cases at least for me I I care a lot less about whether there’s causation there I do care but I care much less about the causation than the raw correlation the reason I’m so interested in the connect is because it says things that have this boast work better or worse so whether that’s the cause of them or not I are happy to imitate the things that do better and not imitate the things that do worse I don’t know whether it’s instantly wrangle you are well aware instantly causation or whether you know it’s a second-order effect or a tertiary consequence or just some some fragment of an effect it doesn’t matter to me I want to look like the people who are successful I want to do what successful beings do and that’s where linkage data is so good for so in part two next week we’re going to talk about some really cool stuff that we’ve knew with correlation data and give you some ideas of where we’re going in the next phases take care everyone you

What's your reaction?

In Love
Not Sure

You may also like

More in:Business