Below is live coverage of the Visual Analytics: Conversion Juice For The Whole Brain panel from the SMX West 2011 conference. This coverage is provided by Keri Morgret of Strike Models.
Disclaimer: The coverage is brought to you in real time, using a custom live blogging tool. Feel free to ask questions or leave comments for inclusion into the live coverage. During the live event, live notes will auto-scroll with newest entries at top. After the session is complete the archive version will have the oldest entries at the top. We ask you to please excuse any typos, as these are live notes.
Finished
Auto-Refresh: Off
Keri Morgret:
10:00:55 am
Visual Analytics: Conversion Juice For The Whole Brain
SEOs are constantly pulling data from analytics programs, keyword research and ranking tools and other tools, but the sheer quantity of numbers can be a challenge to master. A new generation of data visualization tools distill data into images that reveal patterns, trends and other useful information at a glance, sometimes even providing an up-to-the-minute snapshot of website activity by updating in real time. This session looks at a range of visualization tools and how they can aid your search marketing efforts.
Moderator: Chris Sherman, Executive Editor, Search Engine Land
Q&A Moderator: Brian Ussery, CTO, Search Discovery Inc.
Speakers:
Charlotte Bourne, Search Marketing Strategist, Mediative
Wister Walcott, EVP and Co-founder, Marin Software
Keri Morgret:
10:03:40 am
We're starting off with Charlotte from Mediative (formerly enquiro and other companies).
Keri Morgret:
10:05:05 am
Keri Morgret:
10:05:12 am
Objects are websites, relationships are links between them.
Keri Morgret:
10:05:55 am
The graph is an undirected graph. Relationships go both ways. Two-way relationship, like on FB both people are friends. Twitter is one-way at times (I follow you, but you many not follow me).
Keri Morgret:
10:08:10 am
Basic Concepts: Centrality
Three examples:
Degrees: number of incoming or outgoing connections [use for FB status updates]
Closeness: ability to reach all other nodes in the least amount of steps. [use for Twitter updates]
Betweenness: extent that a node lies on the path to reach other nodes. Can act as a gatekeeper by not sharing info between both sides.
Keri Morgret:
10:08:40 am
Wide number of software programs available to provide software analysis.
NodeXL can input a lot of info helpful for SEOs.
Keri Morgret:
10:09:19 am
Keri Morgret:
10:09:29 am
Developed by Microsoft Research, goal is to make doing these charts as easy as a pie chart. Note quite that easy in practice though. :)
Keri Morgret:
10:10:01 am
Keri Morgret:
10:11:28 am
Showing example of client that sells outdoor gear that seven or eight web properties. I'm trying not to take pictures of each and every slide out of respect for the speaker.
Keri Morgret:
10:12:05 am
She grabbed data from Google Webmaster Tools for each of their properties and put it into nodexl.
Keri Morgret:
10:12:34 am
Keri Morgret:
10:14:23 am
With visualizing the links with this data, she's seeing that the Closeout store is getting the most links and the most valuable links, when they'd rather have the main retail site ranking.
Her next steps included looking at analytics for the closeout store and the main site, and see if the closeout store has better conversion rates. If it doesn't, consider emphasizing main retail site.
Keri Morgret:
10:14:58 am
Next example is a competitive analysis, looking at who is linking to competitors but not the client.
Keri Morgret:
10:18:07 am
She's only touched on SEO, but this can be really useful for social media.
Keri Morgret:
10:18:44 am
theresultspeople.com/network
nodexl.codeplex.com/
Keri Morgret:
10:21:13 am
His presentation is going to focus more on paid search.
Simple questions can make a big difference.
1. Is my tracking system capturing data accurately?
2. Do some geographies perform better than others? Should I segment campaigns by geography?
3. Am I giving undue credit to my brand keywords?
4. How long should I wait to assess the value of a visit?
Keri Morgret:
10:22:52 am
When looking at accuracy of conversion data:
They can look at how many third party cookies were dropped where first party cookies were kept.
Chrome: rejection rate only about 32%. iOS worst, rejecting at 80%. Third party cookies rejected by defult in iOS.
Keri Morgret:
10:23:52 am
This can make it look like iOS (iPad, iPhone) traffic converts worse, but in actuality converts better.
Keri Morgret:
10:24:42 am
If more browsers start blocking third-party cookies, you'll want to be sure to use first party cookies.
Keri Morgret:
10:25:40 am
Keri Morgret:
10:25:48 am
How long to wait to assess visit value?
A) Yes, wait for latent conversions, it's more accurate.
B) No, estimate, it's more timely.
Keri Morgret:
10:26:28 am
How long to wait to assess visit value?
A) Yes, wait for latent conversions, it's more accurate.
B) No, estimate, it's more timely.
Keri Morgret:
10:28:16 am
Should I segment by geography?
You can have campaign copies, each targeted to a different geo.
How to see if different geos have different performance. Don't look at traffic, look at conversion rate.
- Web analytics
- Home-grown
-- Pull IP datas for clicks and conversion events
-- Stitch in location data and aggregate (services can help with this)
-- Filter out locations with fewer than 50 clicks for the sample period
-- Put this into a mapping software, such as MapPoint
Keri Morgret:
10:29:03 am
Keri Morgret:
10:30:20 am
Location of brick-and-mortar store had influence on this. If they're further away from a retail store, they buy online.
Keri Morgret:
10:31:08 am
For this client, the suggestion is to try to tie-in information from web visit to in-store sales, such as offer a coupon on web that can be taken to the store.
Keri Morgret:
10:31:32 am
Another suggestion: target only geographies where you do not have a retail presence.
Keri Morgret:
10:32:39 am
Am I giving too much credit to my brand keywords?
To help figure this out:
Tag campaigns into broad categories (easy)
Run both attribution methods (difficulty varies)
Compare first-click vs. last-click conversion counts by campaign and campaign type (easy)
Keri Morgret:
10:33:21 am
Keri Morgret:
10:34:12 am
The result for this company showed they're doing best with branded terms (90% of conversions). he discusses several reasons this may be true.
Keri Morgret:
10:35:45 am
Keri Morgret:
10:36:52 am
This is an interesting but really complicated and difficult to explain in text chart.
Keri Morgret:
10:37:57 am
Interesting to see how many terms are stuck at position two because of quality score.
Keri Morgret:
10:39:18 am
This chart shows where you can spend more money. In this case, non-branded terms.
Keri Morgret:
10:40:18 am
Revenue but not profit maximization technique.
Keri Morgret:
10:40:37 am
He can send us sample spreadsheets behind this if you are interested.
Keri Morgret:
10:40:54 am
A third speaker had been lined up for this panel, but they weren't able to attend.
Keri Morgret:
10:41:14 am
The speaker was from chartbeat, and shows you real-time data.
Keri Morgret:
10:42:34 am
The chartbeat product shows you real-time data about analytics on your site. There are also other companies in this emerging space.
Keri Morgret:
10:42:50 am
Q&A.
Keri Morgret:
10:44:39 am
Charlotte gives example of using the network analysis with Twitter and a hashtag. Look at a hashtag for a topic similar to yours, examine the network, can see how it might behave for your topic.
Keri Morgret:
10:46:44 am
No more questions, so this session has ended.