Why use Google Analytics when you have a Marketing Automation platform with inbuilt analytics capability?

Rod Jacka - Saturday, March, 04, 2017

What’s the difference and why you should consider using both?

Marketers are increasingly asking ‘Why should I use web analytics when I have [Marketo/HubSpot/Pardot/IBM Marketing Cloud]…? Don’t these do the same thing?’
The answer to the first question is yes, you should consider using both.

The answer to the second question is no, they are not the same thing.

Google Analytics is a tool used to analyse the effects of campaign activity, as well as the broader (non-campaign) traffic to your website.

Marketing automation is an excellent tool to use when you have the potential to engage with individual prospective or actual customers and you know who they are. These tools are all about identifying people who are prospective customers and then nurturing them until they purchase and to encourage them to continue to purchase.

Google Analytics is used to optimise and improve the performance of your digital marketing and to analyse the behaviour of all visitors to your website. The data that is used in Google Analytics is aggregated and is in most cases anonymous.

Increasingly tools such as Marketo, HubSpot and others are integrating tagging into how they collect data about people interact with the website. This data is used to help identify and group users of the website into meaningful clusters that can be used to assess engagement and the likelihood that the person will convert into a customer.

Web analytics tools such as Google Analytics collect a much richer set of data from the website and can be used to build a deeper picture about what people do on your website, how your marketing campaigns are performing and where there may be barriers to conversion on your website. Download our whitepaper Marketing Analytics – Your Guide to Campaign Tracking Success for more on that.

The lines between these two types of tools are blurring. The table below will help you identify the strengths of each of these tools and understand how to best use both.

Google Analytics  Marketing Automation
Campaign Tracking Identifies overall campaign performance e.g. conversions from AdWords. Identifies which individual customers have responded to which campaigns.
Identifies areas of improvement from campaigns e.g. bounce rate by landing page. Automatically send personalised follow up email messages to campaign respondents.
Tracks the behaviour of users responding to marketing campaigns on the website. Records which prospects have responded to campaigns.
Conversion Funnel Tracks the progress through the purchase path and identifies barriers Identifies which prospects are in a given part of the purchase funnel.
Tracks where users drop out of the checkout process. Automates the generation and sending of emails to recover abandoned shopping carts.
 Data Collection Data is typically collected from all users and all interactions and then analysed Data is collected from specific actions such as downloading a white paper and then recorded for individual users.
Uses a single JavaScript code applied across all pages to collect data. Uses specific JavaScript code to capture key user interactions. Code generally must be customised for each action to be recorded.
Best Used For Conducting a wide range of analyses to understand user and marketing behaviour to optimise overall website and campaign performance. Understanding where a given individual is within a defined purchasing funnel and to target offers to them based on their preferences.

 

In summary, Google Analytics is a broader and more powerful tool to analyse what is going on with your campaigns and website. The learnings that are generated during the process of analysis are then applied using tools such as marketing automation.

How to Calculate Actual Campaign Profit

Rod Jacka - Thursday, January, 19, 2017

Calculating the actual profit of your campaigns takes time, however identifying unprofitable campaigns is critical.

A robust way to calculate the actual profit is to use the Profit Per Campaign method that requires that the individual Cost of Goods Sold data is available for each item sold.

Where x is the total revenue per item sold from the campaign and c is the total Cost of Goods Sold for items sold from that campaign.

I.e. this is the sum of the total revenue less the total costs for making those sales.

To calculate this requires that each sale has its profit calculated, then an average allowance of advertising expense is apportioned to each sale. A typical formula to do this is:

Where x is the total revenue per item sold from the campaign and c is the total Cost of Goods Sold for items sold from that campaign.

I.e. this is the sum of the total revenue less the total costs for making those sales.

To calculate this requires that each sale has its profit calculated and then an average allowance of advertising expense is apportioned to each sale. A typical formula to do this is:

Profit Per Campaign Formula Detail

E.g. if an item sells for $80 and it cost $50 to purchase this, $5 in handling costs and a campaign generating 100 sales of these products cost $2,000 then the result would be

COGS = $50 + $5 + ($2,000/100)

COGS = $75

The gross profit for this sale is therefore $5.

As the gross profit figure doesn’t include the fixed costs of running a business including rent, salaries, electricity, hosting costs, etc. it seems unlikely that a $5 return on an $80 sale is going to be profitable and hence either reducing the advertising cost for this campaign or increasing the average sale value is required.

Your company’s finance team can probably provide you with the fixed costs for the business or at least a rule of thumb that you can use to estimate these such as for every $1 in revenue the fixed costs are estimated at $0.30.

If the company making the sale above had a fixed cost allocation of $0.30 in the dollar then the campaign would be making a systematic loss for this business.

Fixed cost allowance = $80 * 0.3 = $24

$5 Gross Margin – $24 fixed cost allowance = -$19

The total loss for this campaign is then -$1,900.

It may be possible to estimate that these customers will later spend on additional sales and it is possible to develop models to forecast this. For further information on this type try searching for “discounted cash flow” and “customer lifetime value”.

To help you out we have created a gross profit calculator that you can use to extract the data from Google Analytics and combine this with your cost data to identify profitable campaigns.

How to Calculate Gross Margin Percentage

Rod Jacka -

The Gross Margin Percentage is a key metric that you need to use along side total revenue in your web analytics and reporting tools. The reason for this is that you really need to use profit rather than simple revenue to determine the true value of your campaigns.

The formula for Gross Margin Percentage is:

Gross Profit Margin Formula

Where the value for Cost of Goods Sold includes the purchase price as well as other variable costs such as handling and advertising costs.

The Gross Margin Percentage can then be used alongside the measures that you get from Google Analytics to calculate the total gross profit for the campaign, from which you can then deduct your costs.

In the following screen shot the “B Campaign” generated $82,526 in revenue and 1,067 sales at an average price of $77.34.

Example Campaign

Using the Gross Margin Percentage method we can estimate the profit for this campaign.

Let’s assume for the purposes of this scenario that the average cost to deliver these products is $50 and the allowance for advertising expense is to be a maximum of 20% of gross profit. (Note it is better to work with the actual figures rather than averages, but this makes the illustration simpler).

Our average Gross Margin percentage is therefore:

Gross Profit Margin Formula Detail

Our estimated gross profit for this campaign is therefore:

$82,526 x 28.31% = $23,366.70

Assuming that our costs for the B Campaign were no greater than $5,836. This figure is derived from the allowance for advertising expense of $5.47 per say multiplied by 1,047 sales.

Once we know the full costs to execute the B Campaign once it has concluded or reached a milestone (such as the weekly cycle) we can calculate the profit for this campaign.

Let’s say that the Actual Campaign Costs were $7,500 then the profit would be

Profit=$23,366.70-$7,500+$5,836

= $21,702.70

It is important to understand that there will be rounding and other small errors in this approach to the calculation of profit as it is based on average costs across each product sold in the campaign.

Are Your Campaigns Profitable?

Rod Jacka -

With a little work you can measure revenue and sales per campaign in Google Analytics, Adobe Site Catalyst, Coremetrics, WebTrends and other web analytics tools.

The problem is that revenue and sales are not a good indicator for profitability. For campaigns to be generating value to your organisation they must be profitable which in most cases means that the profit made from immediate sales must be greater than the cost of executing that campaign.

A long-standing measure of campaign performance has been Return on Ad Spend (ROAS). The standard formula for this is:

 

This formula doesn’t however take your actual costs to purchase and deliver that sale into account. Additionally the indicator can be significantly misleading with a 100% ROAS sounding impressive when really it means that your campaign has generated just enough revenue to match the cost of the campaign.

Even worse if your average gross margin on your products or services is 50% then you need at least 200% ROAS. I.e. for every $100 sale it costs you $50 to purchase and deliver that sale. If your campaign costs you $50 to make that sale you need a minimum of $100 in revenue from that campaign not including your other fixed costs such as overheads, salaries, etc.

The result is that in most cases the ROAS measure needs to be very high for profit to be made. Agencies love big numbers and quoting a figure such as 2,000% ROAS sounds great but if you are selling at a low margin you probably need this just to break even.

Solution:

Calculate your gross margin percentage and profit per campaign. These are the figures that you really need to make an informed decision as to the effectiveness of each of your campaigns.

Ideally you will calculate the profit per campaign but this isn’t always possible. In cases where this isn’t possible then the more general approach to calculate your Gross Margin Percentage is appropriate. You can use this figure in conjunction with your Return on Ad Spend and other measures to identify whether a campaign is likely to be profitable.

Where you have the data available then you should use the actual profit per campaign method.

There are two ways that you can do this, the easy way is to use our easy to use campaign profit calculator. For the more mathematically inclined the following is the process to do this manually.

These are outlined in these two articles.

Both of these articles allocate 100% of the sale to the last campaign attribution method. It is possible to use these methods, where multiple campaign attribution allocations are required, by allocating percentages of the profit to each of the campaigns in the attribution model. In this area we encourage you to contact us to discuss your requirements.

One way or another it is critical that you use profit rather than revenue to assess your campaign performance.

Search Marketing – “buy the rumour, sell the fact”

Rod Jacka -

When setting up your paid search campaigns, Google Adwords and Yahoo Search Marketing both suggest keywords to use. This feature makes it easy for you to find popular keywords. It also helps you estimate how many clicks you might get. There are also excellent tools such as WordTracker that can assist you to research keywords.

But popularity isn’t everything; it’s conversion that really counts.

One adage that I have always liked from the world of finance is “buy the rumour and sell the fact”. This means if there is a rumour that a share is about to rise in price, traders will often buy that share based on this rumour. They do this so not to miss out on a potential money making opportunity.

What separates the good traders (i.e. the ones that stay in business) from the poor ones is “sell the fact”. A disciplined trader will rapidly dump any shares they hold when the rumour turns out to be false. They take a small loss and then look for other opportunities. At the core of most successful share trading schemes is a strong discipline to keep losses small and to make the most of any rise in the price.

With a little bit of lateral thinking the same approach can be applied to search marketing. To do this treat each new keyword added to the campaign as a “rumour” and invest some money into them. Once they are attracting clicks to the site then look for whether they convert combined with other performance indicators. Once the “fact” is known a decision can be made whether to sell or change the keyword.

To do this successfully a professional web analytics tool such as Google Analytics or Urchin is needed. You also need to define your website goals and ensure that your campaigns are tagged correctly. With all of this in place, you can then start to assess the value of each keyword. In order of priority here are the factors that I assess for each keyword:

  1. Conversion Rate: Goals achieved / total visits from that keyword. Higher is better
  2. Bounce Rate: Visitors who leave immediately / total visits from that keyword. Lower is generally better
  3. Time on site: How long on average do visitors from this keyword spend on the site. Longer is generally better
  4. Av. Pages per visitor: How many pages on average do visitors from this keyword view on the site. More is generally better

Where a keyword fails to have good values on at least 3 of these it should be dropped. Assessing keywords in this way can help you to priorities your advertising budget and to only take small losses whilst making big wins.

Ecommerce Analytics – A 101 on getting it right

Rod Jacka - Wednesday, January, 18, 2017

Whether you’re starting out on the Ecommerce journey or looking to expand it’s important to remember that your web, mobile and marketing analytics aren’t a set and forget activity. Your use of analytics will change as your business does, and your approach will be dependent upon where you are in the growth phase. We see four main growth phases and outline here what you should be thinking about during each of these.

Just before we head into this, we want to highlight the importance of introducing a framework into your activities and why. A framework is important because analytics can be hard, it’s time consuming and often confusing. You can find yourself lost for days in the detail only to realise you’ve found out a few interesting facts but not actually achieved your goal. Managing change is a methodical process and having an analytics framework is essential. We recommend the following approach starting at the Learn stage:

Analytics is a many faceted area, so at any time you may be learning about what is happening online, offline and in your marketing, identifying friction points, understanding your customers as change occurs, and finding more opportunities for improvement.  A plan is therefore critical so you know what you are aiming to achieve, the timings you are working to and the information you need. You may have a series of hypotheses or ideas on what you think will work or may be the reason why something isn’t working. So you need to test those. To determine whether your tests were successful or not, you need a measurement index to understand where you started from, what your targets were, the end result and then you need to debrief. Whether as part of a team or alone, analyse the results, note what influences were in play at the time and what behaviours occurred because of the change. Then repeat adjusting one criteria at a time so that you are conducting controlled tests.

The above framework applies no matter which stage your business is in. So, now it’s time to look at what’s important to consider during each of those different stages of business growth.

Starting out
If you’re just starting to sell online, getting sales quickly and growing traffic will be your major focus areas. The following are the key metrics you should be looking at and running tests to learn from.

  • Channel growth – are your marketing campaigns and traffic sources delivering enough visitors to your site
  • Quality of traffic – are these sources converting to sales
  • Value for money of campaigns – are you paying too much or is it paying for itself
  • Overall sales
  • Conversion rate per channel

Expanding the niche
In this phase you may find that sales are good, but to expand and grow the business further you need to increase awareness. The following are the key metrics you should be looking at and running tests to learn from.

  • Identifying gaps in traffic sources
  • Relationships between marketing channels
  • Building customer word of mouth
  • Implementing more innovative campaigns to attract attention
  • Measuring social media

Improving Conversion
In this phase you may find that traffic is growing but sales are not growing at the same rate. You will want to understand if there is an impediment to sales and what that may be. Or, you may find that the traffic you’re attracting isn’t the right audience. There could be multiple elements to uncover in this scenario. The following are the key metrics you should be looking at and running tests to learn from.

  • Quality of traffic analysis
  • Product range analysis
  • Conversion rates by campaign
  • Bounce rates on landing pages
  • Friction analysis to determine whether checkout and other key features are efficient
  • Internal promotion campaign analysis
  • Remarketing effectiveness

Powering Growth
You’re in a great position when traffic and sales are good, but this provides an opportunity to grow business profits even further. The following are the key metrics you should be looking at and running tests to learn from.

  • Cross sell and upsell opportunity analysis
  • Site search keyword analysis
  • Customer segmentation
  • Email marketing analysis
  • Remarketing to existing customers

Great, but how do I actually evaluate and analyse the results?

Step 1# Get the data right

The very first thing that you need to do is to ensure the data is right. If you don’t get the correct data, or if your data has errors you can’t make good decisions. We recommend you download the Google Analytics Set-Up Checklist and work your way through.

Tag you’re it! One of the most basic requirements for an effective analytics system is to ensure that you can identify the sources of traffic to your website and how these are performing in generating sales.  If you are unable to identify the sources of traffic then you are unlikely to have enough information on how to optimise these. Bear in mind, if you are using Google Analytics, GA does not recognise all traffic or campaign sources automatically. You need to check your site can track all sources and this may require some manual tagging. A large number of campaign providers automatically tag the campaigns (i.e. AdWords, SalesForce, Campaign Monitor) but many don’t – and Facebook and Twitter are two of those!  A great tool is the Google URL Builder. An important Tip is to set a campaign taxonomy or tracking structure up before you start so that you get a clean, consistent campaign structure.

Another important tip is to reconcile transactions. Missing transactions are credibility killers. We aim to ensure that the difference between actual and recorded transactions in Google Analytics is less than 5%, though this can be higher depending on the site. A few key reasons for missing data may be:

  • Redirections back from payment gateway (e.g. PayPal) not sending data to Google Analytics
  • JavaScript errors
  • Cookie blocking
  • Users exiting the page before the Google Analytics code sends the data

As a final note, Google Analytics is NOT an accounting platform. It needs to have good data but not necessarily all data.

Step 2# Select your approach
There are two major types of analytics uses – the ‘how are we going’ use and the ‘how can we improve’ use. Used together these can be extremely effective.With the ‘how are we going’ use case, you are measuring to see if you are on track to meet your goals. As a team you identify changes in the data and then attempt to explain why this occurred. With the ‘how can we improve’ use case you use data to identify key improvement areas and then implement a structured approach to testing and measurement to see whether your improvements worked. Identifying causal effect is important. Fishbone diagrams, also called Ishikawa diagrams, are an excellent way of brainstorming the possible cause and effect relationships. In this scenario the primary cause may be the number of products, with the secondary cause being the site navigation or search function.

In another scenario, the primary cause for low visitors may be search rankings. Here is an example of a typical analysis path:

  1. Identify which pages are attracting traffic from search engines
  2. Group these by page type (i.e. Home Page, Category Pages, Product Detail Pages)
  3. Identify which pages are underperforming in converting customers
  4. Identify which pages are not attracting traffic from search engines
  5. Estimate what the potential ‘lift’ would be if fixed

Step 3# Build a Routine
Analytics is like exercise, it is best done every day. The following section outlines our recommended approach to building a routine – spreading the tasks out to daily, weekly or monthly.

Daily tasks are all about the ‘How are we going’ question. Each day you should check that the top level metrics are within the expected range.

  • Transactions and revenue
  • Overall sessions to the site
  • Subscriptions or other goals
  • Campaign costs

Compare these to the daily range for each of these metrics.

Weekly tasks are all about short term initiatives and performance. Ask yourself, are you on target?

  • Recap on last week’s performance
  • Document goals for this week
  • Identify actions and required outcomes in advance
  • Note any campaigns that are running
  • Assess underperforming campaigns

Monthly tasks are focused on the bigger picture. Be sure to plan at least one deep dive analysis to be done each month.

  • Month in review
  • Set targets for next month
  • Assess targets from last month
  • Identify tests that you will run
  • Assess tests in progress or completed during the past month

Recap
We’ve covered a lot of ground in this Ecommerce Analytics 101. You should now have the foundation principles to setting up and growing your eCommerce site using analytics.

In closing, we’ll leave you with these key eCommerce Analytics tips:

  1. Make sure you have the Google Analytics Enhanced Ecommerce features enabled and correctly set up – ask us if you need help.
  2. Identify what you need from analytics and how this aligns to your business goals – set up a Measurement Plan
  3. Identify areas for improvement by looking for causal relationships
  4. Build a daily, weekly and monthly routine

If you want to know more, please contact us. We’re here to help.

How to Interpret Time on Site

Rod Jacka -

How long a visitor stays on your website has long been considered a key indication of how successful that site is in attracting relevant visitors. The theory being that the longer someone spends on the site the more interested they are in what you have to offer.

But is it that simple?

Time on site or visit duration can be an indication of the level of interest or involvement that a visitor has with the website. It is also a good indicator of the success of a campaign or other promotional activity that brings visitors to your website.

Often a visitor may come to the website more than once before they purchase, register or make contact. Equally there may be some lag time (often called latency in web analytics speak) between the time that they first arrive and the time that they convert or purchase.

Almost universally when I work with a client on the issue of time on site the question is raised; well what if I leave my browser open overnight, have a cup of coffee, lunch, etc does that impact how long the web analytics tool shows as having spent time on the site.

This issue actually cuts to the heart of a fundamental problem with web analytics – the notion of what is a “visitor” and the assumptions that we use to determine this.

To understand how time on site works it is important to understand how this metric is calculated in your web analytics software.

The time that a visitor spends on the site is calculated as the difference between the recorded time of their last page (or file) request on the site and their first. There are two important concepts here:

1. All requests are recorded. However when a visitor spends a few minutes on a web page and then clicks on a link to download a PDF extra time may not be reported.

2. In the event that the visitor only views 1 web page there is no record of the time that they leave this page or close their browser. In this case the difference between their first and last page request is 0.

As time on site doesn’t account for the time spent reading (or ignoring) the last page in the session it can’t be said to be an accurate reflection of how visitors actually use the site. So leaving your browser open on a page doesn’t cause the average time on site to increase unless you click on another page. In this case each web analytics tool will have its own rules about when to terminate a session and start a new one. In most cases this is based on a gap of 30 minutes or more in the sequence of activities. So if you left your browser open on a page and went to lunch then returned in an hour to click on a link it would be counted as a new session or visit to the site.

Time on site is a good indicative and relative measure. For instance if you compare the difference in time on site for each campaign that you run, then you can see how each campaign compares in regards to the time on site that results.

E.g.
Campaign 1: 2 minutes 30 seconds
Campaign 2: 4 minutes 5 seconds

If the average for all visitors to the site was 3 minutes 20 seconds we could reasonably infer that visitors from campaign 2 where more interested in the site than visitors in campaign 1. Also as campaign 1 had a lower time on site than the overall average, we could say that these visitors were less interested than the general visitors to the site.

As an absolute value time on site is just a number and one that should be treated carefully.

An additional factor is that time on site should be assessed as the average number of pages viewed during the session. In some sites such as online banking you are looking for lower time on site but the number of page views may be high depending on the tasks that are completed. In other sites like a newspaper you would be looking for high time on site with high numbers of page views as an indication of success as the revenue is tied to page impressions.

This is best illustrated in the following grid:

Page Views
High Low
Time on Site High A B
Low C D

The 4 quadrants are explained below. The selection of the appropriate quadrant will depend on both the site and the business goals associated with it.

Quadrant A: High time on site, high number of page views

Positive:
May indicate a high level of interest and involvement with the site.
Negative:
Could indicate high level of frustrations with visitors really having difficulties on the site.
Good Indicator for:

Where a high level of involvement is a key performance indicator. E.g. advertising based sites selling banners based on page impressions.

Poor Indicator for:

Where a site is supporting visitors who should be able to readily access information or perform a task in a short period of time. E.g. government and self service websites like online banking.

Quadrant B: High time on site, low number of page views

Positive:
May indicate a reading behaviour pattern.
Negative:
Lower number of page impressions may be a negative for advertising related websites
Good Indicator for:

Sites that require a lot of time to read and understand the contents of the site. E.g. a professional services company

Poor Indicator for:

Sites that sell advertising

Quadrant C: Low time on site, high number of page views

Positive:
May indicate success for sites that require visitors to complete tasks quickly
Negative:
Could indicate that visitors are lost in the site.
Good Indicator for:

Sites that require high involvement in short bursts. E.g. banking sites, online applications.

Poor Indicator for:

Complex websites such as those of government agencies.

Quadrant D: Low time on site, low number of page views

Positive: For sites that only provide a simple response or quick answers Negative:
Generally implies disinterest in the site
Good Indicator for:

Sites where visitors are seeking answers. E.g. search engines, directories, dictionaries, etc.

In this case repeat visitor behaviour is a crucial contributing factor and must be assessed as well.

Poor Indicator for:

Most websites.

Beyond this there are many other layers that need to be carefully interpreted to understand the value of time on site, however in general it is a good indicator of performance. For further details and assistance with understanding your website please contact Panalysis for a no obligation discussion of your requirements.

 

Free Whitepaper: Best Practice Campaign Tracking.

Take the first step towards improving your marketing’s performance.

Learn more about us Download Now

Reduce the bounce rate and boost the bottom line

Rod Jacka - Thursday, January, 19, 2017

Panalysis undertook a study of over 30 websites to identify what an “average” bounce rate would look like. The results showed that most websites can expect that 1 in 2 visitors will leave their site immediately after they arrive.

So what can you do to fix this?

To start you need to know where the visitors who are bouncing are coming from. It may be that your advertising isn’t working as well as it could or that you are potentially wasting opportunities from keywords in search engines or referring websites.

If you are using Google Analytics then it is a relatively straightforward exercise to identify where the bounces are coming from. To do this start by examining your Traffic Sources -> All Traffic Sources report. Sort the report by the “Bounce Rate” column to see the worst performing sources of the bounce rate. It is likely that the top items will have a relatively low number of visitors, so you may need to move forward a few pages in the report to see the campaigns that attract more visitors but have a high bounce rate.

It is also likely that your site will have a higher bounce rate for certain search keywords over others. Another very useful report in Google Analytics is the Traffic Sources -> Keywords report. Again, sort this by the bounce rate to see the worst performing keywords. As there can be a great many keywords that have a 100% bounce rate and only one visitor, exporting this report to Excel and using this to view the data may be more effective.

The purpose of this exercise is to identify where the worst performing visitors are coming from. Once you know this you can then start to analyse which sources of these visitors are important to your business. It is important to understand that some visitors will turn up on your website and have absolutely no interest in what you have to offer. It is best to leave these visitors alone and do your best to try not to pay for them to arrive at your site.

The second stage is to identify what elements you can control. If you find that your visitors from paid search campaigns have a higher than average bounce rate then you can tweak the campaign to test what elements can reduce the bounce rate. The things that you can change are:

  1. The copy shown in the advertisement that is displayed against a search query
  2. Ensure that the copy and images on the landing page match the advertisement that the visitor saw
  3. Ensure that you tell the visitor in no uncertain terms what it is that you want them to do next.

A similar approach applies for your “organic” keywords. Whilst you can’t control the copy and advertisement that the visitor sees, you can control what happens after they get to the website. The magical thing is that the visitor’s web browser will tell your web server what the address of page where the visitor clicked on the link and hence the keywords that the visitor used. You web developer may be able to assist you to extract this information in real time and then allow you to customise the message shown to the visitor. If your developer is unable to assist, then please contact us because we can.

For instance a visitor may have arrived at the web page of “Big E-Tailer Inc” after searching for “great green widgets”. They were shown a link to the home page because it was optimised for this keyword. When they view that site there is a higher chance that they will leave the site immediately unless there is a big, bold and absolutely clear message that you have all the deals on “great green widgets”. Showing the visitor a generic home page with other products isn’t going to be anywhere near as effective as showing them a page specifically with offers on “great green widgets”.

Using a tool like Google Analytics, WebTrends, Omniture or Coremetrics will tell you what your visitors are asking for, it is up to you how you respond. In general if you are asked a specific question, then respond with a specific answer. Don’t do what most politicians do and avoid answering the question.

The key to reducing the bounce rate is to respond in the most appropriate way to the visitor’s needs. By ensuring that more visitors move to the next stage in the sales process, then it is more likely than not that the overall conversion rate of the website will improve as well.

Web Analytics Frameworks – Who Needs Them?

Rod Jacka -

We all do.

Why? For starters few of us have the time to spend performing in-depth analysis of our marketing activities. However quick reviews of the data result in shallow insights.

A web analytics framework combines set performance indicators for your marketing and website, a sound measurement methodology and processes to assess and diagnose fluctuations in performance.

A well-designed web analytics framework removes a lot of the guesswork from your analysis. It helps your organisation standardise its approach to tracking, analysing and optimising its digital marketing. It also importantly ensures that your organisation agrees on naming conventions and metrics.
In this article I have outlined the five key reasons why you need a web analytics framework.

Web Analytics Tools

But doesn’t Google Analytics (or Omniture, WebTrends, Coremetrics, …) give me everything I need?

No. Web analytics vendors are continuously adding features and modifications to their systems that inevitably increase complexity. They are also trying to solve the reporting needs of a large number of customers. The result is that you have a large number of tools at your disposal but no plan on how to use them together to build something significant.

A web analytics framework is the blueprint against which you use these tools to gather information in the context of your organisational goals and apply this to improve results.

Consistency

Web analytics is not easy. It requires a deep knowledge of the subject, lots of experience and significant skills to perform well. It is even more difficult to consistently deliver results when staff change.

Documentation of processes and training staff in their usage is a key step that few organisations make. It is critical however when creating performance reports that the formulas and data used are absolutely consistent over time. Implementing a well documented web analytics framework is important to ensure that this occurs. Occasionally in our experience with clients, we have seen how a client’s lack of documentation combined with the absence of a key individual can pose a real and significant threat to their business.

Reduced Complexity and Increased Efficiency

Most web analytics tools have a large number of separate reports that measure virtually any aspect of your website. Identifying which of these are important and in what circumstance is the first step to streamlining the web analytics process.

A web analytics framework provides you with a structured approach to rapidly and efficiently track the performance of your website and digital marketing strategy. Such a framework can greatly assist in reducing the complexity of the reporting and mitigating the risk of misinterpreting the data.

Forward Looking Performance Indicators

More importantly a well designed web analytics framework helps you identify potential issues that may impact upon future sales and profit.

For a large number of websites a visitor is unlikely to convert quickly or the sales process may consist of multiple stages, e.g. a membership-based site where a free trial is offered. In these cases a web analytics framework provides forward-looking performance indicators that can be used to predict future sales. For instance if the site offers a 30 day trial subscription, knowing the average ratio of visitors to purchase and their time to purchase is valuable. From this information you can then forecast future sales based on the current number of trial subscribers.

Alternatively if most customers convert a week after their initial visit then any decrease in the volume of visits to your site may forecast decreased sales in the following week. Proactively monitoring search traffic and identifying this drop before a decrease in sales gives you time to take corrective action.

Knowing that your forward-looking performance indicators are not being reached leaves you time to take action before profitability suffers.

Organisational Alignment

An effective web measurement strategy is one that aligns with your organisational goals. Whilst in some cases a website has a clear objective, e.g. to generate leads or to sell a product, for larger organisations there can be separate and conflicting goals. Additionally in large teams there may be conflicting key performance indicators between team members. For instance if the key performance indicators for one team member is to increase traffic to the site and another is to increase conversion these two key performance indicators are potentially misaligned. To achieve the monthly targets the first person may increase traffic from visitors who are far less likely to convert. The second person is therefore unable to meet their monthly targets due to the influx of unqualified visits reducing the conversion rate.

A web analytics framework can alleviate some of this problem by documenting the relationship between potentially conflicting goals and how the key performance indicators can be set accordingly.

Documented Diagnostic Processes

Sales have dropped this month and you need to find out why? A web analytics framework provides you with a process to identify the likely causes of the drop and to suggest a list of actions to take.

In conclusion for most serious websites it is important that you implement a web analytics framework so that you can be sure that you are measuring the right thing at the right time.

Why being statistically correct never hurt anyone…

Rod Jacka -

Businesses are now, thanks to the ease of and access to web analytics tools, using data insights in more and more of their decision making. We’re evaluating traffic sources to understand where our customers are coming from, we’re viewing automated funnels to identify improvement areas in our online shopping processes and we’re measuring and attributing marketing spend to identify the best media options. Some are also running smart hypothesis tests to validate new ideas.

The question is however, are the reports we are using and making decisions from statistically correct? The reality is that in some cases they’re not – and here’s why.

Statistics involves complex problem solving … for a reason.
Statistics for dummies was written for those who want to simplify the complex in order to understand it. It’s a great starter for those who want to understand the basics of statistics, so I highly recommend it (and others like it). The challenge with data interpretation is similar – we try to simplify it in order to make decisions, but this can introduce problems if not done properly. We need to accept that sometimes things are just complex and they require a considered approach to getting the right answer. Learning and using statistical methods to analyse data is necessary for anyone interpreting data and presenting or using results for decision making.

The ‘averageness’ of the average.
A large number of the metrics that are used in web analytics are simple averages – i.e. average time on site, average pages viewed, average sales. Using an average makes the assumption that the distribution of data resembles a “normal distribution” (remember the bell curve). If your data doesn’t match this then the average can be very misleading. We see this day in and day out. The data in web analytics tools like Google Analytics is often taken from distributions that better resemble exponential, gamma or others where averages make little sense.

Take for example the following scenario:

  • 90% of an eCommerce site’s daily sales were be between $119 and $5,271
  • The most common daily sale ranged between $450-$550 per day.
  • However when averaged, daily sales were reported as $1,608 per day

The result: When comparing your daily sales to the yearly average you are going to be disappointed on far more days than you will be happy.

A typical distribution of sales for an e-commerce website.

Data recorded is not necessarily a good measure for analysis.
There are a few good examples of what we mean by this. What we’re seeing in user behaviour today is that we’re multi-device-talented’’- we’re using different devices and browsers simultaneously, but this behaviour is being tracked separately.

Adding to this, not all data that is being recorded into Google Analytics, and other tools is human in origin. Robots and other automated applications can skew the data significantly.

Don’t even get me started on time series analysis. However, this is another important factor as often the data that is analysed in web analytics changes over time – seasonality and other external factors can have a significant influence on short and long term results.

These examples demonstrate one main point: some people using today’s user-friendly analytics tools may not be analysing your company data using statistically accurate methods and great opportunities may be being missed.

Accurate reporting is critical in helping businesses make better decisions on where to place spend and where to focus efforts. By integrating statistics-led-thinking into your reporting, you’ll get a more disciplined approach to analysis, better formed hypothesis testing, and reduced errors arising from data analysis. The good news is this can be done quickly and relatively easily.

If you want help to ensure you’re getting your reporting right, connect with us. The odds are you’ll be better off.

Get the right answers.

Ask us how. Analytics is what we know, and all we do.

Learn more about us Request a consultation

Search