And we’re back at it. I hope you guys had lunch. I, again, did not. I meant to this time, really! Oh well.
[I take back my no lunch comment. Someone just dropped off snacks. Yey, Twizzlers!
Oh yeah, okay, the session.
Seth is up first.
Top 4 Things We Think Are True But Aren't
- Every keyword is equally important (also not true: every person is equally important.)
- All potential keywords are accounted for and known (also not true: all potential suitors are accounted for and known)
- Movement outside the first page does not matter (I'll stop...)
- Education = buy in
Metrics bridge the gap. Chasing rankings is counter productive - but they provide very valuable insight. Traffic/sales is your goal, but rankings are part of the critical path. Rankings are a great way to measure pre-traffic progress. Rankings are a good way to see how you're performing against your competitors.
Call to action snippet optimization
Take 30 days. Every day go to Google at the same time and pull a piece of ranking. Then, take your traffic down for that keyword (like make a note of it, don't kill your traffic). Get some averages. Put your new snippet in place. Then do the same thing for the following 30 days. That will help you find what's the most frequent ranking and you can see how much you've improved your CTR for that particular keyword.
Using paid/natural search percent as discovery mechanism
- Issue: We waste a lot of time trying to determine feasibility on keywords that can easily be tested.
- Solution: Use paid keywords
Destination and volume modified movement
[If anyone can explain to me what this is, I'll have your babies. Seth is clearly way smarter than I am. Maybe Rae or Rhea will explain it to me later. I love you all. Thank you for your time.]
Rand is next.
Rand’s so funny. He looks like he’s about to participate in his high school’s debate. Or give a sermon. Either one. Cute little Rand.
No more guesswork. It’s time for data-driven metrics.
- 30 sites’ traffic data collected via Quantcast
- Comparison with traffic estimates from Alexa and Compete
- Calculated Pearson’s co-efficient correlation
- Confidence ~95 percent
Estimating traffic in the US – Use Compete.com’s Unique People. It’s about .85 correlation. I tried to Google that and found nothing. Again, feel free to make me smarter in the comments. Quantcast came in a close second to Compete.
Global comparisons: Alexa’s 3 month reach is best. Expect data to be 80 percent correlated.
[FWIW, Twitter did not seem to agree with Rand on this.]
Using Third Party Data to Predict Search Engine Rankings
- 109 unique SERP analyzed
- Google Search Results Only
- Correlation between each single metric and ranking position
- Two different data points
- confidence levels of prediction
Importance of single SEO metrics in Google Search, in descending order
- domains linking to URL
- Google.com links
- Yahoo SE External links
- Google.com Toolbar PageRank
- External mozRank
- Links to URL
- Quantcast.com unique visitors
- compete.com unique visitors
Don’t use or trust a single metric to be the key when talking about metrics. When metrics get mashed, expect better results.
What factors predict Google rankings? Mostly like popularity, then content.
Next up is Kelly.
Use many different analytic solutions to get a better story. You want to be able to separate out your traffic and break it down to conversion.
- What traffic are you getting? How much are you getting month over month, seasonally or year over year?
- Where is it coming from? Does it align with market share?
- Growing the slice of pie? Know what moves the needle and what’s affecting search traffic. When did that spike in Google come? What is it tied to?
Keywords: Align keyword traffic with ranking. Find opportunities, missed chances. New terms. Branded vs nonbranded. Has there been a growth in nonbranded?
Site pages: Do your entry pages align with your keywords? What are top post landing pages? Are you missing opportunities there? Are they landing on pages you’d want them on?
Site engagement: How long do they stay? How many pages are visited? What is the abandonment rate? Where do they go? What leads to a conversion?
Referral sites: What other sites drive traffic? Links? Opportunity to change anchor text?
If you have internal search on your site, make sure your analytics package can give you information on that. It’ll give you keyword optimization opportunities. You should also be able to look at location. Where are people coming into your site from? Should you be targeting geographic or local terms? Should you be blocking traffic from certain areas?
Additional Metrics: What are the engines doing? When do they visit? Which pages are crawled vs indexed?
Next up is Brian.
Stop talking about SEO tactics like URL rewriting, titles and meta tags, H1s and body copy, link building, PageRank. Instead, talk about how these tactics will drive business. Start talking Performance Metrics like market opportunity, click through rate, acquisition costs, keyword coverage, etc.
Market Opportunity: Now you can let the data do the talking.
Create Keyword list
- Export your nonbrand phrases from analytics
- Or spice up site to extract relevant on page keywords
Estimate market demand
- Feed into Google AdWords Keyword tool
- Use “exact match” on comparable time period
- Don’t use WordTracker or Keyword Discovery for this
- Calculate your CTR as a benchmark: Divide keyword size by keyword traffic
- Compare CTR by SERP placement
- Model traffic/sales growth potential. Google tactics will increase CTR. Monitor YoY or MoM
Traffic Acquisition Cost
- Compare PPC cost vs natural
- Use Google’s average CPC for each phrase
- Factor in your cost or resource time
Landing Page Yield
- Determine available landing pages
- Pages crawled by important bots
Pages yielding search traffic
- # and % of unique pages
- Indicates optimization quality
- Identify nonperforming pages
- YoY, MoM delta
- Measures the breadth of presence + brand reach. Focus on nonbrand phrases
- Trafficked keywords volume and yield trends.
Incremental traffic and revenue
- More than just “current” minus “previous”
- YoY calculation should consider delta in keyword volume, placement, traffic and sales
- Proper attribution
Return on Ad Spend
- Incremental revenue over incremental costs: Measuring incrementally is key to positive ROAS
- How you get resources allocated: How do you get paid what you’re worth?
- Look beyond short term revenue: conversion rate variance, acquision profitability, new to file customers and searcher lifetime visits