Data Driven Approach
I believe that true marketing is 90% science and 10% sexy, and this is where the importance of data comes in. Being able to gather and interpret large amounts of user experience data is the true key to business success.
Over the last 17 years I have taken a data first approach, whether it is based on ROI, number of new clients, or even website metrics like page views and bounce rate, to really understand my target audience, as well as where to find them.
Data First Approach
I currently find myself doing a lot of interviews for various roles, whether they relate to product or marketing roles, and a lot of the questions relate to “what would you do to increase leads / make better products / drive growth?”. The biggest problem I have with answering these types of questions is my lack of data for the business I am interviewing with. Therefore, my answer always starts with “I would look into the existing data…”.
There is a specific reason for this, it relates to the fact that most companies already have the data to start taking the first steps in meeting their KPIs and goals. What is not always obvious to these companies is, as long as they have been gathering the data, it can be dissected and used to build out distinct client segments, product improvement roadmaps and marketing channel focus.
So where do I start when taking a data driven approach? That depends entirely on what the role requires.
Marketing Data
When evaluating marketing data we look at the 4 categories below:
Ad campaign data:
- Impressions
- Clicks
- Click Through Rate (CTR)
- CPM (Cost Per Mille or Cost Per 1,000 impressions)
- CPC (Cost Per Click)
- Website Visitors
- Dropoff Rate (difference between clicks in campaigns and website visitors)
- Number of Registrations / Conversions (It is important to note, some companies count registrations as their conversions, whilst others have additional steps after registration to consider a conversion).
- Conversion Rate (Number of Conversions / Number of Visitors)
- Cost per Registration / Conversion (CPL / CPA)
- Total Conversion value
- Return on Investment (ROI) (Total Conversion Value / Total Marketing Spend)
Email Campaign Data
- Total Sent
- Number of Opens
- Open Rate (Number of Opens / Total Sent *100)
- Engagement Rate (Not all ESPs provide this, but if they do, it is great data to have)
- Number of Clicks
- Click Through Rate (CTR) (Number of Clicks / Number of Opens *100)
- Number of Conversions
- Open to Conversion Rate (Number of Conversions / Number of Opens *100)
- Click to Conversion Rate (Number of Conversions / Number of Clicks *100)
- Number of Unsubscribes
- Number of Spam Reports
- Number of Bounced Emails (Both hard and soft bounces)
Organic search traffic data:
- Impressions by Search Term
- Impressions by Page
- Clicks by Search Term
- Clicks by Page
- Click Through Rate by Search Term
- Click Through Rate by Page
Earnt Social Media Data:
- Post Views
- Post Engagements
- Post Shares
- Post Comments
- Post Clicks / Website Visitors
- Page Followers / Likes
Each one of these data sets are important and can give you insights into which of your activities are working and which ones are not. Some examples of what is not working:
Paid Campaigns:
Low click through rate
(high impressions, low clicks) – This can be caused by one of 2 things:
- Unrelated ad copy – Your ad copy does not match the keywords you are targeting
- Irrelevant Audience – Your audience targeting is not optimised for your ad offering
High drop-off rate
This is either down to the traffic source quality (Cheap traffic from random sites) or can be an issue with your website tracking.
Low conversion rate
This is affected by both of the issues above, so therefore, you need to be sure that you have evaluated your low CTR or high drop-off rate before you start digging here. If you are sure that your ads, audience and channels are correct, then you need to start looking into your user journey.
This is where you can used advance analytical tools like heat maps to ascertain what customers do on your landing page. Things to evaluate are:
- What is the landing page you are sending them to?
- Is there a clear call to action on the page?
- Does the page guide them through the registration / conversion process?
- Are they leaving your site immediately (Bouncing)?
- Are they moving to other pages of your site? What are these pages? Can the information provided on these pages be added to your landing page?
hese are your 3 biggest data points to look at, as they will determine the amount and the quality of the traffic you are paying for.
Email Campaigns:
Low total sent and high bounce rate
This indicates that the data on your list may be old, or obtained through a 3rd party, rather than people opting-in for your content. To resolve this, you will need to clean your list
Low open rate
This indicates that your subject line is not hitting the right spot with your audience. It could also indicate email blindness from sending too many emails to the same people. To resolve this, ensure that you are not over sending emails to your database, and make sure that your subject line makes people want to open your email.
Low engagement / click rate
This indicates that your message / offer may not be going to the right audience. The low click rate can also be down to the call to action in the email being disconnected from the message you are presenting.
High unsubscribes and spam reports
Again, this is about list quality and the relevance of the message you are sending to this list. You will need to examine where these subscribers come from and determine if you have lost subscribers, or just people who are on your list because they were incentivised to join.
Organic Traffic:
Low click through rate (Page / Search Term)
This means your site is getting ranked well in search, but people are not clicking through to see more. In this case, you need to look at the words people are using to find your page, and see if your meta title and description match the intent of the search words used. Updating these to meet the searcher intent can lead to a significant increase in organic traffic you receive.
Earnt Social Media:
High page followers, low post engagement
This can be down to either having the wrong audience on your page (buying likes) or posting irrelevant content. Having a Facebook / Instagram / Twitter account with loads of followers is nice, but having a high number does not guarantee success. Also check your posting times against when your audience is most active, to ensure that they will see your content. This applies to all of the social media metrics for engagement, as well as your visitors to your website.
Product Data
For product data the approach is a little bit different. There are automated data points that you can use for yo0ur evaluation, however, a large portion of data gathering and analysis for product is down to research and user / customer feedback.
For Product Management we look at:
Usability data:
To gather usability data, the starting point is to look at the actual use of the product itself. Drawing from page usage to actual activation date, you need to explore the entire customer journey to find: Drop-off points, failed use attempts, and roadblocks. This can all be gathered by evaluating statistic that are generated for your product.
Once this data is gathered, you then need to evaluate each of these points as a person using it, to see if there are any obvious issues with the products. After this evaluation, and if you cannot see anything that is not working how it should, your next step is to start reaching out to clients to ask where their issues are.
Engagement data:
Similar to usability, engagement is all about how long people should be using your product, and how that compares to how long they are actually using the product. The easiest way to determine how long someone should be using the product for on a daily / weekly / monthly basis is to use it yourself to see how long it takes you to use the product and then double or triple this time. Remember here, you have built this product, so you know it inside-out. The reason for multiplying your time is to adjust for the learning curve of using your product.
If you have a very high engagement rate, but the amount of throughput by users is the same, you need to evaluate if the product is not too complicated for the basic user. The best way to do this is to reach out to clients and ask them about their experience.
If you have a very low engagement rate, you need to ensure that people are getting their work done, and the outcome of this work matches what you are looking for. If both of clients are getting the work done and getting the desired output, then congratulations, you have built a well optimised product. Obviously, if they are not getting their work done, or not getting the desired output, then you need to investigate why. First by testing the system yourself, and then by speaking to clients about their experience.
Fit-for-use data:
Fit-for-use is a pretty simple thing to check, are people able to add the expected information into your product and get the desired results? If yes, then your product is pretty much fit-for-use. But why would I include such an obvious data point? Simple, cause your product could be fit for use, but it could also be better. Fit-for-use data starts with gathering usage data for the product itself. Once you have gathered that, you can determine if people are getting what they need out of it.
After that, you start talking to clients, and ask them what is missing or could make their lives easier. Remember, although you are responsible for the product and its continued operation, you are not the target user, so engaging with target users can give you great insights on how to get more out of your product offering.
Performance and speed data:
Performance and speed data is again split into system generated data and user feedback. System generated data will give you an idea of how long things take users to complete. User feedback will tell you where your users are being held back from performing better due to the speed of your system.
This is absolutely the task of the product manager to evaluate and raise to the tech team, but unlike most of the other data points on this list, there is very little a product owner can do, other than add optimisation tasks to the backlog (Unless you are also a sysadmin or developer and can also make changes at a server / code level).
This list of data points is by no means exhaustive, and there are plenty of other points you can look at and investigate to determine how well your marketing campaigns or product are doing.
For me these are the bare minimum that you should be looking at, and in regards to optimisation of campaigns or continuous improvement for products, these are very much your starting points.
As you continue down your path of either digital marketing or product management, you will learn that each camapign, every product type will generate its own set of data, which you can then use to maximise your end results.
Get in Touch with me
Use the contact details or form below to reach out and get in touch.
Contact
Email: jweaver14@gmail.com
Phone: 07456392093
