This was my first time to Search Leeds and it was great, so many fantastic speakers, I wrote a blog post of some of the sessions I attended. There have been some thorough write ups already and encourage anyone who did not go to Leeds to read all of them.
Analytics Tracking or How I learned to Stop Worrying and Love Google Tag Manager
Emma Barnes was first up and she gave a cracking presentation from her own experience of Google Tag Manager (GTM).
Emma said that Google Analytics (GA) out of the box is very good. Emma said that with GTM you can get so much from it. So many things, like click tracking, custom dimensions. She said the best way to get started is Pageview tracking. Emma said we need to Select the tag Google Analytics and set to all pages.
You may need to amend your triggers so they only fire if relevant cookies have been accepted (thanks GDPR) as you can see in the photo below:
Emma suggest we also create an event tracking tag. Event tracking is when someone comes to your site and carries out an action besides viewing a page. So you need to know what you want to track. What are they doing? What data do they should I pull.
To start with. make variables to collect data you need. A variable is the bit of info about your website.
Eg you want to know what links they click when they are on my home page.
So this variable is “Click URL hostname ” and the variable type is auto event variable.
Here are some cool things you can do with GTM:
You can see who downloads files, clicks. You can also track c;licks on .pdf .mailto or tel. You can call this category download.
Customer Dimension
Customer Dimension – local time of day (if have website that is multi national, might notice in GA, that you have different spikes throughout the day. By default GA only tells you info about your time zone.
Create a variable to split the data to get local time of day. Create a big of JS for this.
Grabbing author data with customer dimensions, then you can see how your blog posts are working. You need to create a custom author dimension. Put this information on the data layer on the website.
Internal Site Search
How to track internal site search using Google Tag Manager
If someone comes to your site and searches for something, you might want to know what keywords had results and what did not have results. See blow how you would configure this in GTM:
Specific Areas of The Site
- Can you track who clicks a specific are of the site?
- Yes you can. Branded 3 work with a kitchen brand and you can change the type of kitchen.
- Kitchen client wanted to know what areas of the site their clients clicked on.
- Ok this part, you do need the developer to help you. Make sure that you track clicks on specific items.
- Track clicks on “element’s title”
What about SEO?
- You can inject HTML/JS if you don’t have access to your clients website
- But it is not future proof. Sometimes Google see it as cloaking, sometimes does not.
What about Paid Search?
- In GTM there are so many tags that related to online media.
- You can select a tag and then put in an ID and page you want it to fire on.
- Facebook though you cannot. But you can use custom HTML if integration does not exist.
- So Create tag, input ID/HTML
Closing remarks from Emma – Tag manager might be scary but you can do it !!!
Now I feel even more motivated to learn about GTM and implement my tags.
Below is a link to the presentation:
Making Your Marketing Work Harder
Jill Quick gave an excellent presentation especially as the screen was not working so she was talking and keeping the audience engaged when we could not see her slides.
Jill has done 157 audits, ranging from small business to big businesses so she certainly knows what she is talking about. Jill is recommending some easy fixes. You just don’t know what you do not know. This is true.
She also said that Google Analytics (GA) is a ball ache to learn, also true. It takes a long time to learn all the different areas and then Google changes it.
I really like the quote Jill used “without data your just another person with an option”. This is actually by W. Edwards Deming who is a Data Scientist.
There are four different areas of GA
- Acquisition
- Audience
- Behavior
- Conversion ( or you could think of this as Do I still have a job?)
A website is like a house, you throw a party and want to invite people to the party, eg some people from organic, some paid people. Google has something called the Core Reporting API .
Core reporting API , this is like the bouncer on door of website. If you have not tagged that traffic then the bouncer will not have put your traffic in GA.
What are campaign tags?
- They are tiny bits of info, tiny parameters, put on the end of the URLs.
- Mediums = big broad buckets like organic.
- Campaigns are campaigns eg August 2018. It is the name of the campaigns.
- Source = where does your link live. Organic is a medium, the link lives in Google, yahoo,
Then have the URL and the content is the CTA, Now look at the triangle that Jill went through.
- When go into your reports, GA does source then medium. That is why people get the links wrong.
- So Email is the medium. Where does the link live, the newsletter for example.
- If you are not tagging things on FB, Twitter wherever putting the social content. It will get put on the referral.
- But if you are doing paid social media, then you need to tag it. Your paid stuff will be thrown in with the referral stuff.
- If tag it as medium equals CPC, then it goes into your ppc data.
- Other culprits where data goes into the wrong pit is where you have links from http site to the https, will go into the direct bucket,
- Start thinking of other channels where getting inks from the webdste. Eg pdfs sent out, no parameters.
- Staff signatures, google sheets, have traffic, but need to do tagging.
- Go into your GA account, go into all channels.
- Then add a secondary dimension. It is adding another layer of data to your report.
- Type in source and medium, then will give you a list
- Other in GA is the counter does not know where it is coming from. Email campaigns and paid social might be in here.
When go into admin settings, default channel group settings, the checklist is case sensitive.
- System defined, no groupings for paid social.
- If you go into Google demo account, they have removed this from the demo account.
- So amend your settings. Email campaigns, need to have medium email lower case.
- Do this in a test view.
- Test view is a staging environment. On view settings, see channel grouping.
Set up your tagging properly
- If doing paid social and don’t tag it, then it goes into referral. If you tag it as cpc then adwords gets the credit for it.
- Let’s take an example for Search Leeds, if they boosted a post on twitter and paid for that traffic, they want to know where the traffic came from (paid social traffic, organic traffic) so they will need to create a user defined channel grouping called medium= paid social, the source is where the link lives , so that would be twitter.
- Let’s take an example for a pdf from Search Leeds you want customers to read. Again the medium needs to be created and called medium = pdf. When it comes to the source that is for where does the link live?
- For this example it could be Search leeds Agenda.
- It will only happen from the day you created the user defined buckets but from the day you create, you get your own bucket to see what is going o
Clarity within the teams
- Be careful with the agreeing the naming conventions.
- Keep a record of everything.
- Bake it into the SLA with your agencies.
Attribution
- 7 attribution models in GA.
- Google has decided to select in the acquisiton report, it goes to the last non direct channel
- Multi channel funnel report – assisted conversions, email brought in 10 sales, but assisted in 50. This showing that if you are not tagging, you are undervaluiing and over valuing tracking.
- Creating those user defined buckets, that is done at admin level.
- In MCF channel grouping, to copy the grouping and rinse and repeat what did before.
So to do list
- Agree on all mediums doing.
- Go through system defined check list.
- What are we doing?
- Then agree what calling your sources eg newsletters
- Create channel groupings, do in trust view, then test then roll over to reporting views.
- V important for paid social.
- Agree on naming courses
- Keep a track of this
- Think of how can use SLA
Great presentation, thanks Jill. Below is a link to the actual presentation. Jill also has really useful tips on her site: https://thecoloringindepartment.com/resources/
5 False Assumptions About Your traffic
Andraz Stalec from Red Orbit kicked off the third session by saying that search can be rewarding. Based on all that we know on user behavior, we wrote the PPC ads and link to a specific landing page and deliver great results.
But sometimes things do not work out so well. Some surprising things may not expect and it is not your fault.
We assume users do not lie, but sometimes they change their mind.
Most failures are not caused by PPC challenges but it is because we make bad assumptions.
1) Stickyness
- Create a brand stickyness report . For example when looking at a brand like Nike, we only see users acquitted by Nike – used nike terms.
- Adgroup Nike, then see the terms people userd. Eg some used adidas terms.
User this report to do dynamic remarketing, research low stickiness and establish better RLSA.
2) Misconceptions
People think that most loyal customers come from direct. But this is not true. Direct Channel seems to get more traffic. It is quite common to see channels like paid search , organic search and email outperforming direct.
3) Focus on the people who did not convert, this is 98.7% of users
- Use sessions quality report to deliver more actionable insights.
- Direct traffic does not have highest session quality.
- Direct traffic has one of the lowers session quality report.
4) GA attributes all unknown sources to direct traffic.
- Loyal customer may be visiting the site directly but not wanting to buy.But more often than to products and brands, habits are connected to our behavior.
- So hard to tell which part of behavior is habitual and which are not.
- Cross Channel CDJ , First interaction is Direct
- Usually, channel selection is not habitual. You should be omni present and use remarketing.
- Build across Cross channel consumer decision journey report – put on the red-orbit blog.
Read this blog post on Red Orbit to help with your reports.
5) Generic search is in rapid decline.
- Almost 80% drop in paid search revenue.
- There are 3 different types of channel.
- Free fall, brand taking over and generics rule (increase in generic)
- (Market maturity stages and search ) insert the graph.Begging of the market it is genertic. Brand is at the end.
- It is very important to know the types of search users.
- THERE ARE 12 different users for the search query.
You can segment users on 2 dimensions – how familiar they are with recent buyers. (maybe use the slide – 12 types of search users.
Main problem is that we all assume.
- Users searching for a specific brand will by that brand )not true)
- Direct traffic will bring more qualified user (not true)
- Onece you go direct you will alsoway go direc
- Search is cominationed by generic keywords = it is true in early stages of the market.
- You can define user intent from search query. Not true, there are 12 different types of user and you need to understand what user is comong to your website.
You need to know the business, will help make better decisions
- Establish top level business KPIs and connect your activities to them.
- CES stands for client engagement score.
- Create dashboards for day to day monitoring so can have a top view of your business.
- User monitoring system to track your organic traffic to correlated it to conversions, for example SEOMonitor.
- Go beyond traffic, analyse the business. CRM data, transaction data from offline conversions.
Are we getting more loyal users?
- Don’t build reports, create value.
- Offer insights, write down lessons learned and based on that offer actions.
- Prepare some business impact on this.
Very good presentation. Lots of great insights, this is the link to the full presentation:
Content Marketing Tips That Won’t Break the Bank
Kirsty Hulse from Many Minds shared her Content Marketing Tips. She said many brands can get it wrong. For example the campaign from Gregg’s (the bakery) got so much coverage. BUT there was no page for people to go to, to link to and find more information. Such great coverage but then no where to direct it.
Kodak did a campaign, took a video camera out in London and charged people’s phone. Filmed the reaction of them faking wiping people’s data.
All of these did not rely on execution.
We can start thinking brand stories and brand messaging.
Brand stories and brand messaging, are not trusted to take this brand message (to those outside of the PR team) and run with it.
We need to tell brand stories and messages
This is what you need to know about a brand
- What are the 3 core objectives?
- What is the value proposition?
- What is the USP?
- What is the consistent message?
What is important, is the volume.
Kirsty spoke from experience that a campaign she worked on had flopped. Kirsty took client expectations and increased them
- Value questions (what value are you providing me?) will continue to get harder.
- Primary way to mitigate risk is to stop putting all eggs in one basket.
- So run smaller campaigns
- This creates a lot more room for error and for change.
Kirsty explained a campaign that cost less than 250 USD
- Slides is a program, but it does requires basic knowledge of HTML. You can create stunning visual pieces of content.
- Outgrow is another example, they do pub quizzes.
- GET OTHER COMPANIES TO CREATE CONTENT FOR YOU.
- Timeline is another platform to use – can do a lot with this.
The industry is changing, that is fine but we need to be aware and diversify what we are doing.
Kirsty’s slides are on slideshare for a more detailed overview of her talk.
What Happens When a Werewolf Bites a Goldfish?
Sharing Case Studies
- Another content piece they did was look and see what had not been done in the market.
- For example, no analysis had been done for actors, directors or film studios.
- Verve used Box Office Mojo data to create a list of the 100 biggest grossing movies since 1980.
- Then they cross referenced this with IMDb to get the budget. Using both data sets they were able to make ROI calculations.
- The journalists though did not understand the maths.
- They had been outreaching for 3 weeks and no one took the piece of content.
- So James at Verve Search reworked the pitch again and came up with a new title and gave the piece context and contrast.
- What was the result? 120 journalists linked back. It is important to never give up with your outreach, it is not working, try a different angle.
What makes a Hollywood hit?
- Some people say for gaming clients you need to pay for it. It is harder but if giving something valuable to a journalist they are happy to link.
- all data pieces are risky.
- Outreaching a piece can be very difficult.
- Think about your email pitch. Hannah’s colleague had written more than 100 emails and he did not get any response.
- He changed his pitch and journalists starting answering.
Where do you get your ideas from?
- You can get them from anywhere.
- The main difference between writers and other people is that other people realise they are generating the idea.
- You can get your ideas from asking “What if” .
- Often ideas come from two things coming together that have not come together before.
How do you get links?
- Understanding the media landscape is invaluable because breadth of appeal increases your chance of success. You need to give journalists something that allows them to write the stories they want to tell.
- Also pay attention, ideas can come from anywhere, from other smart people, from data, by chance.
- Work hard and don’t give up !
Great tips from Hannah, her slide is also on slideshare for others to read:
Complete Content: A New Model to Drive SEO Success and Much Much More
Creating Knock Out Site Content By Simply Understanding Your Consumers
Craig Rayner from Sistrix kicked off the afternoon session at the Sistrix stage and welcomed David Freeman.
David has been in the agency land for 10 years. He has had a lot of success about some global and national brands. Everyone in SEO is directly or indirectly related to content. Without content we cannot do anything. It is crucial to think beyond SEO as just diving traffic.
- Content is for attracting, engaging and then converting customers.
- And that link building is not the answer to every performance challenge.
- Take people from the search engine to the site to conversions.
Do this across devices. Eg voice search. If your content fails at any of these stages in the photo below, your consumers are lost
- Attract
- Device
- Engage
- Content
How do we optimize content for search engines. The question should be how we optimize for consumers. They are looking for a product and an answer. You need to know what they want.
OK so HOW?
- Create or re shape your on site content strategy
- Adapt to your consumers needs, they expect it. (Scott Abel)
- There is no shortage of remarkable ideas, whats missing is the will to execute them (Seth Godin)
- It is not that we need more content, we need more relevant content (Jason miller)
- Be the best answer (Lee Odden)
SO need to think of this is a multi faceted way and tailored towards differing products/site sections as you can see below:
- Intent is v important. If not thinking about this then the content will not meet the needs of the customer.
- Build personas/ picture of your consumers
- Focus on the funnel to understand the consumer journey. It is not linear
- And consumers content needs at each stage is different as David highlights below:
Augment this with additional data to map content gaps and requirements:
- Paid search data, you get a lot of information from this
- On site search data. People are searching for things
- Google Keyword Planner
- Ahrefs content explorer – you can search based on themes, very good, shows you the top performing content
- Competitor research
Identify the optimal forma for your content – Google it – see what competitors are doing
- Most people don’t do this. The format can be the most critical thing here.
- Create content based on intent rather than search query as consumer intent powers todays search results.
- Add appropriate calls to action aligned to that intent. Ensure the next steop is clear and easy to action.
- Why? Content is there to get customers.
Don’t forget the basics
- A great page title, meta description, and H1 can make a big difference.
- David says he did some tests and made a big difference.
- Conversion rate increased 31% and CRT 6.7% for the UK Treatwell.
- Makes a big impact.
- Finally make sure the relevant technical optimisation is present.
- This approach propelled Kia to No1 in organic search market
- Kia was the first car company to build category pages, eg SUVs,
- They placed reviews front and enter. No link building required.
- Nationwide also gained No1 organic market share vs direct competitors following these principles:
- Titles and meta description optimised using paid search insight
- Consumer focus IA and innovation to surface the most relevant content
- Editorial focus on guides and life stage content.
- No link building
Summary
- Discover the content needs across all stages of the content journey from the search engine through to conversion
- Identify the optimal format for your content
- Create the content for the consumer not search engine
- Test content to maximize traffic and conversion rate.
Thanks David, great tips and here are David’s slides for you all.
How to Optimise the Sh** out of your internal Search
Luke Carthy is the Digital Lead at Mayflex and shared his experience about internal search.
He said by doing an audit of the internal search, they had 4 million search URLs removed and meant a 30% increase in MoM organic traffic on the Mayflex site.
What internal SEO is essential? There were 15 reasons but Luke said 3 (there wasn’t enough time to go into detail about the others).
- Can guarantee countless new URLs
- Keyword cannibilization
- Impacts on conversion/UX
Below is the screengrab.
Category search
There are so many filters and so many possible URLs in one category,
Real Life Examples – Argos and Tool station
Luke gave us some super juicy examples how bad SEO has been done.
Argos
- Argos has 4.5 million pages in Google’s index, 85,000 products, 101,000 search urls. They have more search URLs than their products. That is insane.
- By using Ahrefs, Luke showed us that can see terms of search, more back links and domains besides revenue to the search page. But revenue is the browse URL.
- Site:argors.co.uk/search “smart tv”
- Then Luke showed us the category URls vs Search URLs. Category URls were the smallest but made the biggest revenue.
Tools stations
- 16,000 products and 39,000 search URLs
- So almost half of their site is search urls. URLs that had no real purposed.
- Tootlstation.com/search and toolstatation.com/shop (this one has much more revenue)
- Site: toolstation.com toolbox – put this into Google and then see there is duplication content.
- 50% of their site is search urls but all search URLs drive less than 5% of sales and barely any traffic.
So what options do you have?
No one size fits all solution. But plan of attack can use:
Step 1
- Size up your search URL issue use Deep Crawl and Google Search Console.
Step 2
- Identify the link juice to search URLs
- Use ahrefs for example,
Step 3
- Then when identify them, 301 search urls with worthy backlinks.
- Also identify search URls with good organic traffic and generate sales.
- Prepare to lose/de-index all other search URls.
- <Meta name = “robots” content =”nofollow, noindex”>
- Don’t forget to nofollow, noindex search filter URLs too.
- Change out internal links pointing to search URLs.
- So then what has worked for Luke is a combination of nofollow noindex and redirecting
Final tips
- Keep tabs on the organic traffic levels
- Check in on number of URls in index (you want this to fall)
- Keep close eye on core KPI factors like revenue etc
- If alternate landing page doesn’t exist, make one
- Refrain from adding in a robots.tx condition to wipe out search queries. Makes hard to identify queries with internal search.
- Don’t be afraid to delete the millions of URLs, you have nofollow
- Monitor future backlinks and actions.
Luke’s presentation on slideshare for everyone to read:
Conflicting Website Signals and Confused Search Engines
Rachel Costello from DeepCrawl gave an in depth presentation about Technical SEO. Rachel looked into how we can take power back from Google as they choose what they show.
Not all website owners have knowledge of SEO, imagine what their technical implementation looks like.
Google uses an algorithm for URL seletion for those who do not how to proper canonicalise. You can sit back and relax and Google will choose the URL, but the URL is not perfect.
Tweet from John Muller – don’t confuse Google unnecessarily.
- How Google handles website signals and selects URL
- How you can rest the signals your website
- Sfd(photo)
Subscribe to DeepCrawl webmaster tools hang out.
The canonical tag Is a signal not a directive.
If you have two pages with new content, once will ignored if you use canonical tags.
Does Google process canonical tags in the rendered version or raw HTML? They said they do not (process in the rendered version)
6 Website Signals
- canonicals
- Internal linking
- Sitemaps
- Parameter handling
- Redirects
- Backlinks
All feeds in that the content is not duplicate.
- John Muller said in the hang out that parameter handlings signals are stronger than the canonical signal.
- With website signals, there is a strength in numbers. Make sure everything is aligned.
How can you test these? Which elements to test?
- Internal linking
- Redirects
- No index
- Sitemap inclusion
- Disallow
Use a tool to get this information. Eg Rach used DeepCrawl.
Examine non-indexable pages being shown – Rach recommended this one.
Why would these URls be in the search even though you said not to?
- Could be due to internal linking
- Google may choose a redirect URL instead of the target
- Don’t use no index on canonicalised pages – do not use a non index on canonicalised pages cause if Google accidentally picks a noindex page as the canonical, nothing with show up in search results.
All signals need to tell the same story to get content indexed by Google
How to test Google’s URL Selection:
- You can see which canonical Google has chose in the new Google Search Console.
- Also have pages with proper canonical tag (look at GSC)
- Also go into the SERPs and see it for yourself
- Use the info query and the cache for this.
- Info:www.deepcralwing/xdfsfdd
- Info query only shows canonical URls.
Sum up
- Understand Google’s URls selction algorgithing (canonical listaion, internal links)
- Monitor your site for conflcting signals
How to Audit Your Site for Security
Julia Logan, aka Irish Wonder shared her experience on hacked sites and what you can do if your site has been targeted. Julia started off by saying that any pharmacy related queries will have a bunch of unrelated sites in them. They have been hacked. But you wont see warnings the sites have been hacked.
- It’s Google Webmaster Guidelines that say you are responsible for monitoring your site’s security and hacked content – hence can lead to penalties
- You are the only person responsible for your site’s security
- Some people say that with SSL your site is secure but that is not correct.
Why do sites get hacked?
It is nothing to do with you or your site. Any site can get hacked, but most of the time it is something on a scale. Not targetted at one site specifically.
- People want to create parasite pages on established domains to gain the authority of those sites
- Hacking for the purposes of crypto mining. Infecting them with malware then uses their system for crypto mining.
- Top CMS usage stats, wordpress is responsible for running more than half of existing sites with datasource. So high risk of being hacked.
More vulnerabilities has been reported last year than ever before.
Two scenarios of a security audit:
- Ideal scenario is a planned health check when going through, checking vulnerabilities.
- Fire alarm – when already been hacked and need to do something v quick and sort it out.
Planned hacked scenario should be part of your site audit.
Check what gets indexed for your site
- Do you want plugin directories indexed? This may lead to hackers targeting your site.
- Do you put out any sensitive information that might get indexed? How many web development agencies keep xl files of their clients’ details.
Check for known Vulnerabilities
- Plugins and platforms using that are known of existing vulnerabilities. May want to use cve details to help.
- Remove any unused plugins.
- Check if you give anyone access to your system without any particular need.
Eg copyrighters do not need admin access.
Check if your version is the latest version
WPScan is what Irish wonders recommends for wordpress.
Security by itself is great but it can only see so much from the outside. But that applies for any tool doing external checks.
Check your SSL certificate – use this tool Qualys SSL Labs to check.
Your SSL implementation is only secure if you consistently link to secure resources.
What about the fire alarm scenario?
- Do a site search for your site and add spam keywords to see if you have any parasite pages.
- Or you might have been hacked for another reason, so this would not work.
- If your site suddenly gets much slower than maybe something is going on on your server.
- Power tip – check your server logs to see any unusualo URLs being requested. Check Majestic for your indexed /linked to pages
- Check Google Search Console for unusual queries, URLs and crawl errors
- Make sure you have a clean back up please
- Do not delete anything untul you know what happened.
- Update all passwords – ftp password, cpanel
- Remove vulnerable elements that may be guilty if there is no fix
- If there is an existing fix, then update the vulnerable element.
- Check if your mail server has been affected.
Julia Logan, aka Irish Wonder’s slides are below:
How to Fix the Most Common Technical SEO Issues
Craig Campbell gave an indepth presentation about common SEO issues we all face and showed how we can fix them.
He said the first thing we need are good auditing tools. It used to be that SEMRush was amazing for keyword research, but it is also vey good at website auditing and pdf reports and sending them to trello. SEM Rush can also create great dashboards.
- Craig went through one example of Coca Cola.
- They had a score of 44% once put through the SEMRush tool.
- When do SEMAudit, can see SSL and crawlability, errors and warnings.
Basics of an audit:
- Improved on page seo image tagging, missing meta data, internal link s
- Remove any errors and problems.
Extras
- Do a site speed and page load and time to first byte.
- Do not think you have to have100% for your site.
- Also data and testing,
- SEMRush has done a test on many websites (150,000 random sites) they found that 82% had issues and 44% had critical issues, so there is a lot of data within SEMRush
SiteSpeed
Google doesn’t like slow websites
- Look at GT Metrix
- Pingdom
- WebPagetest.org
Trim down page size, videos, images, javascript and CSS.
Tools you can use to do this:
- JSMin (javascript)
- YUI Compressor
- Closure Comiler
- Gzip/Deflate Compressions
- WordPress Plugins
Some tools
- Wp Fastest Cash
- SEMRush
- NJLazy Load
- Speed Booster Pack
One reason for slow load speed can be large HTMP page size.
Common Servicer issues;
- Crap hosting
- 6High traffic websites
- No browser caching
- Excessive plugin use
- 25% with no caching
- Too many redirects
Cloud flare – lots of benefits. Helps make site faster. There is a free version as well.
- Also HTTP2
- Free SSL as well
- TTFB – time it takes the server to get the first byte from the server.
- Time to First Bite
- Reasons for slower TTFB – network issues, lots of traffic, RAM usage, Data base, shared server.
- Ways to improve TTFB
- Caching, load small files
Some great tips for us to implement, more details are in Craig’s presentation: