In the last couple of years, I have been asked numerous times for a “dashboard”. In reality I was asked a chart showing “what is my project (‘s KPI) doing”. I usually performed a few data pukes and provided one PowerPoint slide, once the KPI is clearly defined, and the source data is in a “reasonable-messy” format. I was not impressed by the business reality of over-used term “dashboard”.
“What is a dashboard”
In my opinion, a true “dashboard” should serve beyond “what”, but also “why”. To achieve that, a well-designed dashboard needs to be “interactive”. For example, “drilling down” into at least one more layer to understand detail, and “slicing” it by date to provide context and benchmark for comparison purpose. The reason is simple – a “what” questions is being answered by a report, and a dashboard should serve the data in a more aggregated level with better flexibility – better equipped to answer “why”, and ultimately driving “actions”.
I’d like share some of my learning from a recent digital dashboard project, from both
- business side – how and why recommending certain metrics and KPIs in the digital world,
- technical side – building it from handling integration of multiple data sources, to telling a compelling data story using Microsoft PowerBI.
The focus of this blog post will be the business side.
“Economic value as the #1 metric”
It is fun working with a great marketer to brainstorm and building it together -The process includes interviewing marketers on the marketing plan, bouncing ideas on what shall we measure and why, and even poking thru all different digital touch points to get a “visitor point of view”. The goal is to determine the appropriate metrics & KPIs to truly reflect the effectiveness of digital marketing effort. At the end of the day, they should be all leading to business impact, which ended up being only one metric: Economic value $$ to the business(credit to Avinash’s awesome blog post).
In the actual implementation of his model, I find it is a much harder sell to the business than I initially thought, especially if the actual value is showing a much smaller number than the investment. The alignment with the marketer is key here, and he/she is the one who provides key assumptions as they are closer to the business. Also, a detailed metric breakdown to the economic value is necessary, to boost the confidence of the key business stakeholder. The desired outcome is making the business believe “Despite our current economic value might be low, we won’t use it to blame anyone. Instead, we will use it as a key input to guide the success of the digital evolution”
In the following paragraphs, I’d like to break it down into three main digital space: Content, Email and website.
For simplicity reasons, the social media (rented platform) and blog (owned platform) metrics have been grouped together as the same “content marketing”. At the end of the day, they are serving similar purpose – generating demand for the top half of the funnel. After relentless prioritization, the following 5 KPIs are recommended:
- The number of social media post: Instead of using vanity metrics “impressions” or “reach”, we choose this one as it could indicate a high-level quantity of social media appearance. It will also be using as a denominator for other KPIs.
- Engagement per post: For consistency purpose, we use Share + Like + Comments per post across social platform as the key indication for the level of effectiveness of social media post. Using this one, multiply by the first one, we can easily get the total engagement the brand in the social media channel.
- Owned blog incoming traffic: The absolute number is a key indicator of the next level funnel conversion. However, the breakdown of it is more actionable. For example, if only 2% of the total traffic is coming from organic search, SEO effort is clearly needed.
- Owned blog conversions: The conversion is determined by the business goal of the blog posts. It could be a light-weight micro-conversion like video watched (if it is embedded), or a heavy-weight macro-conversion like leads sign up(if a form is presented).
- Blog traffic driving to the company’s website: Not many visitors will walk down this path – the industry average is only 2.3%. We’d better take care of them if they are.
The email piece is standard across the industry. It could also shift significantly, with the growing popular trend of shifting to marketing automation.
- The number of emails delivered: A high-level quantity metric, using a denominator.
- Unique open rate: an indicator of the email being opened by the audience. The actual result might be higher than the reported metric – the challenge of one many email clients won’t allow downloading picture by default. If the picture is not rendering, it won’t count as “open”.
- Unique click thru rate: A more accurate metric than open rate, but usually the total click thru is too small to make it “statistically significant” to make decisions.
- Unsubscribed rate: showing a negative action taken by audiences.
- Email traffic driving to company website: similar to content marketing piece above
Company’s website is usually being considered the “backbone” of the digital journey, as the “real conversion” is mostly made within the website, usually in the format of lead generation or eCommerce transaction. At the same time, the website tracking technology is more mature comparing to the other two, providing a lot more metrics options, and higher data confidence. For consistency purpose, I have followed a similar pattern of “funnel thru” to recommend the following KPIs.
- Traffic to the sites, and its breakdown by traffic source. Similar to traffic to owned blog, the breakdown can offer much more actionable insights.
- Resource download: Grouped all micro-conversions into this “bucket”, the purpose of this metric is to provide business a sense of how the digital resources, including white-paper, brochures, video content, are being consumed. Ideally, we will be able to make optimization based on its breakdown.
- Business outcome: Usually it is a combination of “leads generated”, and “buy now” traffic – no matter it is eCommerce transaction or directed to distributors.
- Economic value: Simply using a dollar assumption for each of the KPI previously, excluding the traffic. For example, 200 downloaded of white paper X $5, plus 5 leads X $300, plus 400 buy now clicks X $15, will give us an economic value of $8.2K. The challenge will be to come up with the assumption dollar value.
A dashboard is another data puke without
- Insights and recommendation delivered by the analyst
- Active involvement – interpreting, questioning, brainstorming by the dashboard consumers.
Instead of taking a “hands-off job done” attitude, I will recommend to schedule follow up meetings with the marketers and business stakeholder to review the dashboard on a regular basis to
- Discuss context, and come up with insights and action together
- Getting 1st hand feedbacks to fine tune KPIs and dashboard.