Statistics, R, and You: Advice for a New Analyst?
-
Hey SEOMozers!
Two prongs to this question; I'll keep it succinct.
I've been working as an in-house SEO/SEM Analyst for about 5 months now. While I'm generally savvy at telling the story behind the traffic/conversion data, and making forensic recommendations (I worked in SEO prior to this while in college), ideally I'd like to see my reports read less like these piddly Excel charts and percent change statistics. Ideally they'd look more like Nate Silver's FiveThirtyEight blog for the New York Times, or OkCupid's periodic dispatches on OkTrends: visual, statistically-informed, and predictive, the kind of report that under other circumstances might plausibly generate backlinks.
Data analysts swear by R for statistical modeling, but is it useful for our Google Analytics data sets, holes and uncertainty and all? Is the steep learning curve worth the effort? Tutorials I've seen online assume a proficiency in programming or statistics that's beyond me, or they're written to support a textbook exercise. Any recommendations for a book, online course, or general resource with more of a niche focus?
And a general question about stats too, since it's related: what level would you prescribe if I really wanted to kick this up a notch? I studied a humanity in college and while it helps with the numerical storytelling, I wonder if the practical arcana of Bayesian Methods/abstract probability theorems have a place in Web Analytics. Do they? Are there options for us bushy-tailed young analysts to pick this up without resorting to B School?
Thanks in advance!
-
What Steve said.
Especially:
"Slipping in one or two suggestions about direction and budget in the verbal reports can be very beneficial."
I've found verbal briefings every week or two are far more effective than written reports. Sometimes we take a quick dive into the raw data in a "guided tour" type of way. Often that goes away over time. But first you have to establish trust. That means getting some wins user your belt. And it also includes fessing up to failed initiatives.
I see a lot of people get hung up on metrics of questionable significance and obsess over reporting. They never ask a basic question: "Is any there evidence my reports are being read?"
Of course, meetings and calls are far more easily arranged with small business owners than in big, bureaucratic organization that lack a single decision-maker. But I'd move heaven and earth to try to get a monthly 20 minute meeting with someone as high up the food chain as possible. Pretty soon, he'll start inviting others and stuff will really happen.
-
Hi,
I've just replaced an "In House" SEO with a consultancy model. The process related to me from the in-house guy was very bureaucratic as you have said.
From a consultant perspective I need to prove what I do does have a difference to te bottom line and then this builds trust.
Work out what will make a difference to the bottom line and attack that first. In my situation, they'd been running AdWords unsuccessfully for a while so I managed to get a small budget (£100) to prove a point. We ran a campaign for 2 weeks and got a noticeable spike in traffic and business. This has built the trust in my approach and process enormously.
On the reporting side I only report once a month in writing with other ad-hoc verbal reports of how things are going. Slipping in one or two suggestions about direction and budget in the verbal reports can be very beneficial.
Report less - Do more.
When reporting try to focus on the "what you did, what outcome you expected and what you learned" This should then feed back into an SEO Plan and overall strategy. In terms of SEO I've found business leaders like strategy.
Hope that helps
Steve
-
Love a good wacky, "out there" response to keep my intuitions in check!
I get what you're saying. That's happening, but in-house SEO/SEM is a slow and measured process, subject to approvals that go far up the chain of command. I figure if I can at least present my recommendations in a more salient way, they'll be able to defend themselves whether or not I'm in the room.
We also, as with most hotels, deal with affiliate sites. The cost/benefit of involvement there has considerable room for interpretation, and while it isn't exactly my onus to understand it, I'd like to. Basically looking to grow here.
-
Right.... make a change on the site and show how it increased sales.
Show what you have done that has hit the bottom line.
-
Ready for a really wacky, "out there" response?
Stop writing so many reports and start trying to influence decision makers in other ways.
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
More 'New users' than 'Users' in Google Analytics
Hey Mozzers, I'm seeing more 'New users' than 'Users' in Google Analytics (see screenshot). This is over a 3 year time period. Any idea why new users would be greater than users? Thanks! 6lYmU0j
Reporting & Analytics | | RWesley0 -
New Site Worries
To cut a long story short, our old web developers who built us a bespoke site decided that they could no longer offer us support so we decided to move our back end to the latest Magento 2 software and move over to https with a new company. The new setup has been live for 3 weeks, I have checked in webmaster tools and it says we have 4 pages indexed, if I type in site:https://www.mydomain.com/ we have 6560 pages indexed, our robots.txt file looks like this:Sitemap: https://www.mydomain.com/sitemap.xml Sitemap: https://www.mydomain.com/sitemaps/sitemap_default.xml I use Website Auditor and Screaming Frog, Website Auditor returns a 302 for my domain and Screaming Frog returns a 403 which means I cannot scan any of these. If I check my domain using an https checking tool some sites return an error but some return a 200.
Reporting & Analytics | | Palmbourne
I have spoken to my new developer and he says everything is fine, in Webmaster tools I can see some redirects from his domain to mine when the site was in testing mode. I am concerned that something is not right as I always check my pages on a regular basis. Can anyone shed any light on this, is it right or am I right to be concerned. Thank you in advance0 -
Are these Search Console crawl errors a major concern to new client site?
We recently (4/1) went live with a new site for a client of ours. The client site was originally Point2 before they made the switch to a template site with Real Estate Webmasters. Now when I look into the Search Console I am getting the following Crawl Errors: 111 Server Errors (photos) 104 Soft 404s (blogs, archives, tags) 6,229 Not Found (listings) I have a few questions. The server errors I know not a lot about so I generally ignore. My main concerns are the 404s and not found. The 404s are mostly tags and blog archives which I wonder if I should leave alone or do 301s for each to /blog. For not found, these are all the previous listings from the IDX. My assumption is these will naturally fall away after some time, as the new ones have already indexed. But I wonder what I should be doing here and which will be affecting me. When we launched the new site there was a large spike in clicks ( 250% increase) which has now tapered off to an average of ~85 clicks versus ~160 at time of launch. Not sure if the Crawl Errors have any effect, I'm guessing not so much right now. I'd appreciate your insights Mozzers!
Reporting & Analytics | | localwork0 -
New e-commerce launched, drop in traffic
One of our clients launched a new e-commerce site and there was a significant (60%) drop in traffic in Google Analytics following the launch. We are about 10 days into the new site, and our suspicion is that the move has caused a decrease in their site authority which has lowered their search position. The old site had the store at sitename.com/store. The new site has a Shopify store at shop.sitename.com. There were very few external followed links to the store pages, so we thought that there would be little effect for overall site traffic by moving the store to a subdomain. Any recommendations for next steps on diagnosing the problem and working towards a solution?
Reporting & Analytics | | DesignHammer0 -
Google Links Disavow - Does that preclude new links from a domain?
If using Google disavow links tool and you disavow links from a 'domain' does that mean that any 'future or new links' from that domain will be blocked? Answer Yes is good if the domain is spammy but bad if the domain was submitted in error ........ Answer NO is good if the domain was submitted in error but bad if the site is spammy. Does anyone have an answer to this please? Also is there a disavow 'undo' request process available? cheers, Mike
Reporting & Analytics | | shags380 -
Need a tool for finding new links from analytics
I'd like to get a weekly report of all of the brand new referral links that have come in to my site over the last week. Is this something that can be done in google analytics? Is there a better tool for this out there? Thanks in advance
Reporting & Analytics | | seo-hunter0 -
If I change the URL of a page, but the old page canonicalizes to the new, do I need to change my Analytics goals to get data?
I changed the URLs of some pages recently (because the same thing that affects the internal anchor text also affects the URL - grr...) but considered it not a big deal because even if I looked at the source code of the old URL, the canonical tag was now pointing to the new one. The question is - if I had URL destination goals set up for those URLs in Google Anlaytics, do I now have to change them? Or does Google somehow know that anyone getting to the new URL is the equivalent of someone getting to the old URL because of the canonical tag that exists on the old URL source code? I still do see goal conversions for some of the old URLs even since I changed them - but it could be that people are still somehow finding the old URL somewhere - or that Google only reindexed it a week or so after I made the change. Any light to shed? Thanks in advance, Aviva B
Reporting & Analytics | | debi_zyx0