So you’ve been handed the keys to a web analytics platform with no context? No company goals, no helpful colleagues, just raw data? It’s a challenge, but not an insurmountable one. Even without a clear roadmap, a couple of hours of focused analysis, guided by your experience, can unearth valuable insights.
Think of it as a quick data-gasm, not the earth-shattering experience you’d get with proper planning (as discussed in my previous post), but a pleasure nonetheless.
Setting Realistic Expectations
Don’t expect to uncover revolutionary insights in a few hours. However, you can identify solid starting points for deeper analysis. These starting points can inform customer-centric strategies (“What are my customers telling me?”), reveal gaps in your online marketing (“Where am I wasting money?”), and help you prioritize your efforts.
A Beginner’s Guide to Succ: 10 Starting Points for a Strong Start!
This guide outlines the steps I typically take when starting with a new website’s data. It reveals where to find valuable starting points, even without specific knowledge of the site, business, or priorities. We’ll cover reports to examine, KPIs to evaluate, and inferences to draw.
Here’s what we’ll explore:
- Visit the website. Note objectives, customer experience, and areas of suckiness.
- Assess the acquisition strategy. (Traffic Sources Report)
- Determine visitor loyalty & recency.
- Identify and fix easily correctable issues. (Top Landing Pages)
- Pinpoint profitable content. ($Index Value Metric)
- Evaluate search strategy sophistication. (Keyword Tag Clouds)
- Distinguish between profitable actions and mere activity. (Goals & Goal Values)
- Optimize the marketing budget. (Campaign Conversions/Outcomes)
- Enhance the experience for ready-to-buy customers. (Funnel Visualization)
- Discover overlooked issues. (Analytics Intelligence)
Ready to dive in? Let’s go!
Step #1: Visit the Website. Note Objectives, Customer Experience, Suckiness.
My biggest gripe with web analysts is their eagerness to jump straight into the analytics platform. Before you drown in data, experience the website as a user.
Visit the product pages, the donation pages, the contact form – whatever is important to the business. Sign up for the newsletter, read customer reviews, download white papers, and use the site search.
alt: Heuristic analysis of the Fry’s Electronics website, noting areas of positive and negative user experience.
Get a feel for the company’s brand, information architecture, user experience, etc. What is awful? What is great?
Bonus points: Visit a competitor’s website and do the same.
Take notes. What did you like? What did you hate? What frustrated you? What was obviously broken? What is the site trying to achieve?
At a minimum, answer these questions: What is the macro-conversion? What are two or three micro-conversions? These apply to both e-commerce and non-e-commerce sites.
The client didn’t provide guidance, but you now have valuable context. You’re ready for the data!
Step #2: How Good is the Acquisition Strategy? Traffic Sources Report.
This is where I start, because I want to understand how effective the company is at online marketing. All other site data is secondary, because poor online marketing limits the potential of even the best website.
Every successful company has a balanced portfolio of acquisition channels (traffic sources). How’s yours?
Look for:
A balanced portfolio of traffic sources: Search, Referring Sites, Direct, Campaigns. Which are strong? Which are missing?
Ideally, you want something resembling the example on the left.
- 40% to 50% Search is normal. Too high indicates over-reliance on search rankings and algorithm changes. Too low means missed opportunities. A significant portion should be Organic, indicating good SEO rather than solely relying on paid search.
- Around 20% Direct Traffic. If implemented correctly, this represents existing customers and offline campaigns. A healthy amount of both is desirable. Low direct traffic raises concerns about customer service and retention.
- 20% to 30% Referring Sites. A healthy web strategy includes traffic from other sites that link to, recommend, or discuss your products and services.
- 10% Campaigns. Email, display ads, social media campaigns, etc. This demonstrates a deliberate effort to attract targeted traffic.
These are general guidelines. Your mileage may vary.
Next steps:
Identify areas where the company is overleveraged. Dig deeper and brainstorm diversification strategies with the client.
Examine the top ten rows of each traffic source. Dig deeper into one or two sources to identify strengths or weaknesses. Look for surprising keywords, missing traffic sources, and trends in campaign vs. direct visits.
You’ll gain an understanding of the client’s sophistication and where to focus your initial analysis.
Step #3: How Strongly Do Visitors Orbit the Website? Visitor Loyalty & Recency.
Now that you understand the site and its traffic sources, it’s time to focus on the visitors.
Many sites lack a clear purpose or have misaligned content and engagement strategies.
You need to understand how attached visitors are to the site. This is crucial for content sites, but also important for e-commerce sites (retention, repeat purchases, support).
The Visitor Loyalty report reveals how often users visit the site in a given period.
alt: Google Analytics visitor loyalty report showing the distribution of visitor frequency over a one-month period.
Look for:
In the first example, 47% are “one-night stands,” but 40% visit more than 9 times a month. This reveals content consumption patterns and helps you prioritize tasks (segment the 40%, understand what works, and apply it to the 47%).
If the data resembles the second example, with mostly “one-night stands,” there might be a problem.
Consider the impact of these loyalty patterns on the objectives you identified in step one. Use this data to drive discussions with the client.
Also, review Visitor Recency, particularly for content sites. This measures the time between visits from the same visitor.
alt: Google Analytics visitor recency report showing the time elapsed between visits.
A high percentage of visitors returning within 24 hours indicates a strong connection.
Segmenting this data by content, source, campaign, and outcomes is beneficial, but even a quick review reveals visitor behavior.
Next Steps:
Review Length of Visit and Depth of Visit.
alt: Google Analytics length of visit report showing the distribution of visit durations.
These reports reveal what visitors do once they arrive. (Remember that visits with one page view, bounces, will be in the 0-10 seconds bucket).
If you have time, segment the larger buckets and analyze the data. If not, knowing Loyalty, Recency, Length, and Depth provides insights into visitor engagement.
Step #4: What Can I Find That Is Broken and Quickly Fixable? Top Landing Pages.
You understand the site, traffic sources, and visitors. Time to get your hands dirty.
Companies invest heavily in traffic acquisition, often inefficiently. Identify landing pages where money is being wasted. Are visitors refusing to click? That’s a useful signal.
This is a standard report in most analytics tools. It displays bounce rates, indexed against the site average, for the top entry points.
alt: Google Analytics top landing pages report showing bounce rates indexed against the site average.
Look for:
The red! The “indexed against site average” feature makes it easy to identify underperforming pages.
In this example, three landing pages perform well, while seven have significantly higher bounce rates. The cause is unknown at this point.
A high bounce rate suggests either:
- The wrong people are arriving (problem with campaigns, SEO, etc.) or
- The page is poorly constructed (missing calls to action, broken).
Without more time, pick two of the worst offenders. Click on the “arrow thingy” to visit the page, then view the page-level report and segment by Entrance Sources and Entrance Keywords. This reveals the sources and keywords contributing to the high bounce rates.
Now, for a few pages, you know they are underperforming and have clues as to why.
(Exception: When analyzing bounce rates for blogs, segment and focus on landing pages for New Visitors.)
Next Steps:
Recommend specific fixes. You’ve reviewed the pages and traffic sources (a proxy for customer intent). Use a heuristic evaluation process to identify improvements that will reduce bounce rates.
A helpful checklist is available here: Qualitative Analytics: Heuristic Evaluations Rock!
Step #5: What Content Makes Us Most Money? $Index Value Metric.
Most website effort is dedicated to content creation. Focus on page reports, but switch from an input metric (bounce rate) to an output metric: $Index value.
For e-commerce sites with revenue, or non-e-commerce sites with defined goal values, $Index value calculates the “revenue” earned (or contributed) by a page. It’s a good gauge of a page’s value.
Go to a content report (e.g., Content By Title) and look at the last column:
alt: Google Analytics content by title report showing the dollar index value for each page.
Look for:
The types of content that contribute most to your website’s goals. In this example, the value added ranges from $8.79 to $0.10 even within the top ten.
Wouldn’t your boss love to know exactly what types of content to create more of? Or which product pages generate the most value?
Also, review Content by Drilldown.
This report is useful for sites organized with clean directory structures (e.g., /products, /videos, /blog). You can identify which content groups are most valuable. Are videos valuable? How about demos? Do wishlists work?
If you lack a clean directory structure, segment your content to achieve the same results.
Next Steps:
Use the Analytics Weighted Sort feature.
alt: Google Analytics report showing the Weighted Sort feature used to prioritize high-value pages with growth potential.
This answers the question: “Ignoring the highest and lowest $Index value pages, which currently high-value $Index pages should I focus on to maximize my bottom line impact?”
This requires algorithmic intelligence in your analytics tool. Clicking that checkbox provides the answer.
Focus not just on current successes, but also on hidden areas with future potential.
Step #6: How Sophisticated Is Their Search Strategy? Keyword Tag Clouds.
Content is covered. Time to return to high-value acquisition: Search.
It’s difficult to gain a “big picture” understanding of search strategy from top ten tables.
Instead, extract the data and create a tag cloud of all 40,000, 50,000, or 100,000 rows. Export as CSV, paste into www.wordle.net. Magic!
alt: Keyword tag cloud for the Church of Jesus Christ of Latter-day Saints, showing prominent keywords related to their search strategy.
Look for:
(The following examples use data from competitive intelligence tools).
A tag cloud quickly reveals the story across hundreds of thousands of keywords. In this example, for The Church of Jesus Christ of Latter-day Saints, the brand dominates (as it should), but many other valuable non-brand words are also prominent. This validates that the words the LDS church wants to be associated with are prominent, and those it doesn’t want are less so. The themes (music, scriptures, family, etc.) align with business goals.
Create these views for Paid Search or Organic keywords to gain further insights.
Tag clouds are great for understanding the strategic picture and the sophistication (or lack thereof) of any brand. Compare the LDS church with Chase Bank:
alt: Keyword tag cloud for Chase Bank, showing the dominant keywords and overall search strategy.
For Chase, analyze the cloud. What do you think?
Even after removing the large words (representing massive traffic) from the cloud (e.g., “Chase” and “Bank”), the story doesn’t improve much.
How about this one:
alt: Keyword tag cloud for Blackberry, illustrating the keywords driving their search traffic and overall search presence.
Simple data presentation, big insights.
Tag clouds have limitations. They don’t pinpoint the root cause. Is it people? Lack of sophistication? Cruise control? Bad SEO? Further investigation is required. However, you now have: 1. A great understanding of the site’s search data and 2. Valuable insights to present.
Next Steps:
I’m a big fan of internal site search analysis. Few sources contain as much direct customer intent. Visitors are directly telling you what they’re looking for. The challenge is aggregating that intent into something understandable.
Download the data and create a tag cloud.
alt: Tag cloud visualization of internal site search queries, highlighting frequently searched terms.
Rather than analyzing page views and Google keywords, you can quickly understand what people are seeking.
Analyze where people start searching, how many bounce from the search results page, and the conversion rate/goal values for the top searches.
Mine the data to identify topics to write about or products to sell.
Step #7: Are They Making Money or Making Noise? Goals & Goal Values.
After a detour into search, we return to outcomes.
I can gauge the sophistication of a business by what I see in the Goal Conversions & Goal Values report:
alt: Google Analytics goals and goal values report displaying key performance indicators and conversion metrics.
Look for:
First, check if you see anything at all.
If the report is empty for an established company, it indicates a struggle ahead.
If macro and micro conversions are present, rejoice!
Verify that the goals & conversions align with your notes from Step 1. If not, what valuable actions are visitors taking? Are there new goals to create? What do the trends over the past 12-16 weeks suggest? Which goals contribute the most value?
You should be able to confidently discuss how the website is meeting business objectives, and where it’s falling short. If you’ve completed Steps 2-6 effectively, you’ll have actionable recommendations.
Next Steps:
Configure any missing goals in the analytics tool.
If behavioral goals are lacking (as they often are), create them, referencing your insights from Step 3 to set the threshold values.
For e-commerce sites, create a segment for “whales” – customers with orders 50% higher than the average order value. Identify their origin (geographies, campaigns, keywords), on-site behavior (content consumed), and purchases (shopping cart analysis).
Understanding these valuable customers is essential.
Step #8: Can the Marketing Budget Be Optimized? Campaign Conversions/Outcomes.
Remember the three important outcomes in web analytics: More Revenue, Reduced Cost, Increased Customer Satisfaction.
This step focuses on reducing cost. Even with limited client input, you can identify opportunities to reduce marketing/customer acquisition costs.
My favorite report: Goals / Conversions by Campaign:
alt: Google Analytics campaign performance report showing campaign conversions and cost per conversion metrics.
Look for:
Campaigns encompass Paid Search, Display, Email, Social Media, and any valuable sources discovered in Step 2.
Start with the “Other” report in Google Analytics (or the “Campaigns” report in your tool). Focus on the “Per Visit Goal Value” column as a measure of efficiency. A large difference ($1.02 to $63 in the example) clearly indicates superior campaign performance.
Then, examine the conversion numbers and the individual goals contributing to the high value.
You should identify potential areas of wasted or inefficient spending and opportunities for improvement.
Next Steps:
Select one or two major campaign strategies and delve into e-commerce analysis (if applicable). The Paid Search and Yahoo! Display campaign examples highlight the potential to uncover significant deltas in conversion rates and average order values.
In web analysis, we often obsess over behavior or acquiring more visits (spending more money!). To gain attention (and approval), initially focus on cost reduction. This lesson comes from experience.
Step #9: Are We Helping the Already Convinced Buyers? Funnel Visualization.
Let’s focus on the quickest way to improve business outcomes.
The underappreciated funnel report. While unstructured path analysis is a waste of time, structured path analysis is invaluable.
alt: Google Analytics goal funnel visualization showing user drop-off at each step of the conversion process.
You want users to progress through a series of steps to achieve a goal. Credit card applications, e-commerce orders, donations, and leads.
The funnel report reveals where people abandon the process.
Look for:
The red bars, especially the biggest ones.
That’s it. If the funnel was correctly created.
Identify the highest exit points in the funnel. Visit the corresponding pages, get feedback, identify improvements (heuristic evaluation is acceptable), and submit them for fixes. Implement the obvious improvements and A/B test the rest.
Imagine someone filling their cart at your supermarket, lining up at the cashier, taking out their wallet, noticing a long line, and leaving. You don’t want them to leave! Fix anything hindering the open wallet.
Next Steps:
If the funnel spans multiple visits (sessions), standard Google Analytics (or Adobe Site Catalyst or WebTrends or NetInsight) funnels won’t work accurately.
Use a tool like PadiTrack. It measures pan-session funnel conversion performance. Visitors can enter, exit, and convert across multiple sessions, and you can track that behavior.
PadiTrack also allows segmented funnels. Search, Display, and Email visitors convert differently, and this tool reveals those variations.
PadiTrack is free, uses the Google Analytics API, and works on historical data.
Step #10: What Are the Unknown Unknowns I Am Blind To? Analytics Intelligence.
Without any guidance, you’ve completed nine steps and found actionable insights.
But the risk of working without context is missing something crucial.
The unknown unknowns!
Before finalizing any analysis, review the Google Analytics Intelligence reports. Google Analytics uses algorithms and statistical analysis to identify anomalies in the data.
alt: Google Analytics intelligence reports with monthly alerts, showcasing automated anomaly detection.
Look for:
Initially, set the Alert Sensitivity to Low (multiple standard deviations from the mean) and review the automatic alerts. These indicate significant events. Then, gradually increase the sensitivity to identify other alerts.
Seek events and activity, on the site or externally, that you (and the client) might be unaware of.
These discoveries can validate or contradict your previous work and impress the client by uncovering insights they missed.
Next Steps:
The challenge is isolating the root cause. Examine the GA Intelligence Major Contributors section for clues. Use advanced segmentation to isolate the source, behavior, and outcome causing the anomaly.
(For more information, watch the videos on Google Analytics Intelligence.)
And you’re done!
Summary: A Beginner’s Guide to Succ: Tips and Best Practices for Web Data Analysis
Here’s a handy checklist:
Step #1: Visit the website. Note objectives, customer experience, suckiness.
Step #2: How good is the acquisition strategy? Traffic Sources Report.
Step #3: How strongly do Visitors orbit the website? Visitor Loyalty & Recency.
Step #4: What can I find that is broken and quickly fixable? Top Landing Pages.
Step #5: What content makes us most money? $Index Value Metric.
Step #6: How Sophisticated Is Their Search Strategy? Keyword Tag Clouds.
Step #7: Are they making money or making noise? Goals & Goal Values.
Step #8: Can the Marketing Budget be optimized? Campaign Conversions/Outcomes.
Step #9: Are we helping the already convinced buyers? Funnel Visualization.
Step #10: What are the unknown unknowns I am blind to? Analytics Intelligence.
The first time you follow these steps, it might take longer than 120 minutes. But with practice, you’ll improve.
Reading this post isn’t enough. You need to practice on many clients. Your sense of direction, data, discovery, and deduction will improve over time.
Ideally, begin with a clearly defined web analytics measurement model. But even without one, you now have a framework for delivering valuable insights from any web analytics tool, for any website.
