Analytics Every SEO Needs To Know - It's more than just rankings and traffic reports to measure the health of SEO efforts. This session focuses on analytics that SEOs should be considering.
Moderator: Rand Fishkin, Co-Founder and CEO, SEOmoz
Q&A Moderator: Matt McGee, the rockin smallbusinesssem.com dude
Brian Klais, Executive Vice President, Search, Netconcepts Laura Lippay, Group Program Manager, Search Strategy, Yahoo Jonah Stein, Founder, ItsTheROI Richard Zwicky, President, Enquisite
The first speaker is Brian Klais. 10 Essential Natural Search KPIs.
There's more to search success than measureing ego term hits and rankings. - How do we evaluate channel performance? - What's the missed opportunity cost? - How do you identify pages and markets to focus on? - How to prioritize further investment - etc. The metrics include: - Brand to non brand ratio, unique pages, missed opportunity cost, ROI and brand reach, page placement, yielding pages, visitors per phase, indexation rate, and more.
Case study data of an anonymous merchant which gets 73% of traffic from Google, 12% from Yahoo, 4% from Google Images, etc.
Brand to non brand ratio - a lot of people get traffic from their own brand name. They're happy with that, but he says that you should really focus on the long tail instead. - Long tail search = 40x brand search - There's remaining opportunity The non brand queries dominate.
How big is your site? - You can't answer that if your site is database driven. Look at how many pages the bots see. Non-duplicate pages crawled by your favorite bot. - Look at the size of your site. That forms the top of the funnel. Think: email sent. It gives color to... Indexation rate: - Pages crawled vs. indexed * Your advertising inventory Why aren't some pages that are crawled not in the index?
Yielding pages is a critical concept. Of all pages in the index, how many drive traffic? How many aren't? - Identify those "free loading pages" - title tags, etc. If you know what pages drive traffic, at what rate are they not driving traffic? How many vistors do they not get? Visitors per phrase roll up into a metric.
Page placement - inequality equals opportunity. Identify those opportunities to get them from page 2 to page 1 becasue you can get a 3x increase in traffic (in this particular situation)
Engine Yield Rate - ROI of the engine's crawl. - Google is returning 3.6 visitors for every page that it crawls. Yahoo returns 0.3 visitors for every pag.e MSN live is 0.1 visitors. ROI and brand reach are critical metrics but it's difficult to calculate the cost of this especially internally - labor cost, expertise, etc. He advocates focusing on brand, don't discount it. You're winning unbranded keyword searchers to your site - how do you measure that?
Missed opportunity cost? How you calculate it - there are 50,000 pages on this site. 20% are generating traffic at a rate of 11 visitors per page. Every order is a 2.8% conversion. It''s $34 per page. Think, though: what more can I get? What if the market size is 5x more? 50000 unique pages x $34 per page - $1.6m/month = $15m/year
Laura Lippay is the most important person at Yahoo. Prove your worth. Do what most SEO's can't. Dazzle your boss by proving how your SEO is better - wow your boss with the grid which is inspired by Bill Hunt's opportunity matrix (Craig Hordlow's balancing paid and organic listings session).
What is it? - Keyword based data - - Balance SEO/PPC and paid inclusion -0 what channels perform best - Referral Gap - where do we have content but aren't performing well - SEO content opportunities - what search terms do we not have content for? - Make SEO traffic and value projections - if we ranked #10, what is the dollar amount?
What do you need? - Use a graph for all these projections. Graph looks like this. She shows a grid of how you can measure this data. - Gather keywords - Add keyword data for performance comparisons - Just see the performance of one channel - With search volume and CTR byposition, make projections. On a big set of terms, when you have this, you can see the worth of it if you put effort into it.
What is the worth of a #5 ranking, or a #2 ranking? You still need to be realistic.
Showcase your skills: add demographic data, ranking data, and use those to create reports.
Jonah Stein is here!!!!!!!!!!! He's cool. - Five forgotten metrics: 1. Customer lifetime value. Most people don't track this. We spend time looking at analytics package but we forget to look at the big picture. About a year ago, Eric Enge and he did an analytics study and they realized that the AdWords conversion tracker was getting more data than ever - becasue of cookies even if ROI is stolen from your SEO campaign. - Stolen ROI: keyword revolution (brand claims ROI on the short tail, long tail, and brand); campaigns steal ROI - email blast, coupons, internal promotions.
What you really need to do is - customer relationship manager - real ROI and LTV. - write permanent cookie on first touch - capture source, keyword, and date - write first touch data to CRM on conversion event - capture missing data at every touch point (order surveys and support interactions like chat, phone, and email)
Once you have that info, you get more accurate LTV and ROI - - Captures cancelations, adjustments, and reorders
Crawl frequency tracker from blogstorm.co.uk - he gets about 100k visitors a month. He shows how much times he's visited by each spider. Crawl frequency turns out to be a meaningful statistic that you can find out on-page. - Toolbar Pagerank is deliberately misleading - Google doesn't want us to know. - Crawl frequency, however, is useful for you. Obfuscating this will lower the quality of the index. - Relative CF is great for diagnostics - shows canonical URL, replaces supplemental index
What governs crawl frequency? - Numer and quality of inbound links - Sitemap settings - Server data - Content update frequency
How often crawler visits, crawl depth, saturation, etc.
Measure crawl rate - seometer.com, crawl rate tracker from Patrick Altoft, log file analysis, custom solutions, etc. It's really a simple PHP script to see when spiders arrive. If you have developer resources, do it yourself.
Pageviews to conversion metric - send them to the right page. When we say the number of pageviews per user, that doesn't mean it's good or bad. It could be that your SEO was done really well (you took them to the right page. Or you abandoned people).
External links - we don't look at how many links we give out. The number of external links per page - it's great to have a wide footprint of pages. As SEOs, we need to pay attention to how many links we're putting out and to what pages.
He had to cut his presentation short but it was very good.
Robert Zwicky is next! He's cool also. I met him in a cab in NYC, I think.
We're all SEOs and our work is undervalued. Chart on the board links to the pages on the SERPs that people click through on your site. He pulls down data with 3 million queries - text based, image based, etc. Referral traffic from page 1 is at 90%. Most people don't go beyond page 1. He says that Rand Fishkin is a traffic whore. Anyway, he wants to figure out how to do better because not many people are signing up for Pro memberships (and you all really should. I wrote a guide for them). Anyway, he wants to know what drives traffic from page 2 to his site. Using the Enquisite data, you can tell where the links need to come from in order to bump up that traffic. Drill in and focus on terms that you want to to drive converting traffic - you'll win every time. You may not have to optimize on page. You need to figure out what's not performing in your target market. Use your analytics to let you win everytime.