How to create effective dashboards and scorecardsDate: Aug 06, 2009
The previous chapters in this guide uncovered dashboard best practices and advice, revealed the latest dashboard and scorecard trends and explained how to use dashboard editors for streamlining and increasing user adoption. In this chapter, SearchDataManagement.com turned to BITadvisors CEO Brian Jordan, who recorded a video screencast that divulges more best practices and design tips for dashboards including:
- Different types of dashboards and what they should be used for
- How to meet dashboard design challenges
- Focusing on usability and how to define the needs of different users
- How to meet business and technology needs through effective design
- The most important metrics and goals to track in your executive dashboards
- Key characteristics of optimized dashboards and scorecards
In the video you'll see several examples of optimized dashboards and scorecards used in the real world and learn what makes them so effective. Also see examples of poor dashboard design and learn what to avoid when creating your dashboard. Finally, Jordan provides seven top tips for successful dashboard and scorecard design.
Don't miss the
other installments in this dashboard guide
How to get started with dashboards
10 key elements for effective dashboard designs
Executive dashboards and data visualization trends and future outlook
Working with dashboard editors for streamlining and increased user adoption
Real-life examples of effective dashboard design
How to create effective dashboards and scorecards
About BITadvisors Inc.
From the initial choice of the most appropriate business intelligence solution to successful implementation to end-user training, BITadvisors is a "one-stop shop" for companies seeking BI software and consulting. At BITadvisors, our team of BI professionals have many years of experience planning, implementing and supporting BI solutions.
Our customers range from small, up-and-coming organizations with a handful of employees to successful Fortune 500 companies. From basic business reporting needs through the most sophisticated analytical applications, BITadvisors has worked with them all. In addition to a wide range of customer sizes, BITadvisors has worked with customers in a variety of industries, including healthcare, financial services, manufacturing, and life sciences. For more information, visit www.bitadvisors.com.
(Editor's note: BITadvisors was acquired in June 2010 by Kogent Corp., a Santa Clara, Calif.-based BI consulting and data management services firm. Brian Jordan became executive vice president of sales at Kogent after the acquisition.)
To learn more about CRM dashboards and other specific CRM and analytics products, visit 20/20software.com.
Read the full text transcript from this video below. Please note the full transcript is
for reference only and may include limited inaccuracies. To suggest a transcript correction,
How to create effective dashboards and scorecards
Brian Jordan: Hello, my name is Brian Jordan. I'm CEO with BIT Advisors. We are a premiere business intelligence professional services company here in New England, and we specialize in data warehousing and business intelligence solutions for our customers. I'm here today to talk about creating effective dashboards and scorecards, and I'm going to talk through how we do that for our customers.
The agenda is to talk a little bit about some of the challenges that we meet with dashboards. Who are the users of dashboards, the different types and categories that we break dashboards into and then best practices on how to design dashboards. So, what are the business users looking for out of a dashboard solution? Very interactive, looking at high-level metrics. You'll hear KPIs, you'll hear scorecards, those types of terms. Really, it's the ability to look at key performance indicators for the business, for the department, for you as an individual, but it's also very intuitive, very interactive, getting a lot of information in one place.
It also has the ability to get some underlying details. So, I might see a high level metric, but I want to be able to drill down to find out what the root cause of the issue is. From an IT perspective, we're looking more at flexibility, ease of development. So, we hear all the time that IT can be a bottleneck in terms of reporting and getting information back to the business.
You really want to minimize that where dashboards are concerned. You want it to be easy to turn around, easy to augment or change. You want to make sure that security is implemented properly. So, as a user I log in. I only see information that's pertinent to my job and decisions I need to make. And then, it's important to be able to leverage any existing investment that you already have. A lot of times our customers already have a BI platform, and the dashboard is just an extension of that, or there are some kind of niche dashboard solutions, but they tend to integrate very nicely with an existing BI platform that you might already own.
In terms of the different dashboards and how we categorize them, they're really based on the needs of the users. Operational dashboards are very detailed. They tend to be updated more than once a day, even it could be up to the minute real time. So, for example, if you're looking at monitoring inventory or how many products you have in inventory, you might want to do that up to the minute in real time.
If I'm a call center operator, I'm taking an order over the phone. I can see the data. We have that available. I can ship that right away. So, very detailed, operational in nature, as I mentioned. Supervisors, line managers, call center people, etc. Monitoring really the operations of the business.
Tactical dashboards are often tracking yourself or a department against the goal. They also are very detailed. They look like an application, so if you go to Salesforce.com, you can log in, you can get a dashboard of how much revenue you've sold for the quarter, and see how you're tracking against your quota. There is typically daily and weekly information, so a little bit more historical in nature. You're reviewing versus looking up to the minute.
Strategic dashboards are, you'll hear executive scorecard, executive dashboard, a cockpit, etc. It's really designed to measure the performance of the business as a whole, key KPIs, key performance indicators. What are the key measures or metrics that drive this business? It's often looked at by executives, senior level management. It's very departmental, enterprise focused summarized information, refreshed, often quarterly or monthly. Those are the three different types, three different categories of dashboards.
Let's talk about how we would design or go about best practices for designing these dashboards.
Key characteristics. It should be high level information. So, you're not going to see lots of detailed data in the dashboards, down to the low level, the transaction level. It's going to be very visual, so you're going to be able to log in and very quickly get a gauge or a chart that tells you how you're doing visually. And it's also, you can even have exception based reporting in there. So, you might log in, and there's an alert that tells me, 'hey, there's a problem here. I can drill into the detail.'
The dashboards are very interactive, very intuitive. They rarely require training, so that's one of the nice things that I like about them is they're very intuitive, and as such, you can log in and click through. It's very easy to understand, and it's very easy to use.
There are also a lot of guided navigational tips, so we try to prevent people, our users, from getting into trouble. One way to do that is to build a navigation. So, if I'm in a high level metric, I click and boom. It brings me the detail that I want to naturally look at. It's in both historic and real time, so we can have trending data over time, or we can have up to the minute data for inventory or things of that nature.
Here's an example. You'll see over here I’ve got a high level total, actual versus budget, and then it's broken up by different sales reps. I can see how they're doing now, what's their budget, or what's their forecast and then the variance. Down here I can see their contribution to the total sales. And then, on the right I've got some information in both real time and historical data. So, I can see how I've been trending over the last six months. I can see my targets here, and then if it's forecasted, how I'll do in the next six months based on this historical information. Those are the key characteristics. Again, easy to use. Very visual.
Here's an example of a best practice dashboard, and you'll notice some things right away. I've got a scroll here. That is maybe, my management or exception based management reports. Increase expenses, I get some text that might be an underlying detailed report. Over here on the left I've got two sets of metrics. I've got sales and cost, so these are my KPIs around sales reporting and cost. I've got profitability growth, sales volume.
Let's look at sales volume. I can see down here over the five month period how I'm doing. I can make that 12 months if I need to. Now, that's great, but I want to see how was I doing versus last year. So, I can click on this last year. I can see, wow, I'm actually doing a little bit better over the last couple of months, but I've taken a little bit of a dip here for July.
Right now I'm looking at all regions, all divisions. I can break it out. Let's just look at North America. It looks like in North America, I'm doing OK. Let's find out Latin America, same thing. Asia Pac It's a little different. Asia Pac, it looks like the reason for the dip in overall sales volume. If I wanted to look at the divisions within there, so you see I can get some analysis done fairly quickly.
At any point I can run a total sales report which will give me all the customer transactions for that region and that division which will maybe, tell me, OK, it was this customer. I can say, ‘hey, it looks like they got a really heavy discount this quarter, and that's why the sales numbers are off.'
Over here I've got a company broken up by the different divisions. I can look at it year-to-date. I can look at it quarter-to-date. I can look at it monthly. It looks like it's tracking pretty well towards the year, the month, but the breakdown of the different divisions looks pretty good, pretty even over the course of the year. So, you can see I'm getting a lot of information presented to me here. I can then switch over to growth. If I want to look at the cost perspective, I can do that. So, lots of analysis, lots of interactivity here on this one dashboard.
So, what a dashboard isn't, it's not a data mining tool. It's not going to give you heavy analysis capabilities, like an ad hoc reporting tool or a data mining capability. It's not going to give you lots of detailed information. That's really enterprise reports. So, we start high level and then allow you to drill into the detail once you get to a key metric that's of interest or an issue. It's not an iteration back and forth with the data. And so, think of an analyst who wants to ask and answer multiple questions. It's not that either, but it will and it should integrate easily with some of your other capabilities, some of your other platforms that you have in house.
In terms of, like I said, what it isn't, you can see up on the top here I've got more of a data mining tool. Down here this is an enterprise report with lots of detailed transaction, detailed information. And then, on the left I've got some ad hoc capability. So, back and forth with the data. You can see I'm pulling it in a year, revenue. I ask it that. It triggers another question, so I go back, like an ad hoc query. That's not a dashboard.
So, in terms of good and bad design, this is really what I would call a good design best practice. You'll see I've got year-to-date total, total average, and then I break it into the regions over here. And then, down below here I've got a trend so I can look at it monthly, yearly or weekly, and the chart changes on the fly. A bad design over here is where I'm just getting way too much information.
I think of a dashboard as a suitable replacement for 10 or 15 different reports. So, that's what this is trying to do over here on the right. It's just presenting too much information. I've got high level down here. Then, instead of having the radio buttons over here, I've got a monthly, yearly and weekly all in one place. It's hard to look at, and why would you want to look at it that way. I'd rather look at it yearly and say 'OK, that's interesting' and break it into quarterly, 'oh, that's interesting' and then break it into monthly. So, that sense of drill down.
Lastly, we've got some best practices and tool tips for success in terms of designing and implementing dashboards. I think the best approach would be a proof of concept. Let's take something, some low hanging fruit within the organization, get some business buy-in for it and create a proof of concept, a dashboard or two, and then figure out how the whole thing works, how the implementation should go.
Then, based on that, with the approval and the sign off with the business, you're going to design out the rest of your corporate departmental business dashboards with that design in mind.
Scalability is important so you might start out in the department, but long-term, you're probably going to want to roll this out to the entire organization. So, you want to choose something that you can scale, that allows you to look at a lot of the different data across multiple business processes.
Don't over visualize. Don't over design. You saw that example of a bad design where you are just putting too much information on one page. It's OK to have multiple dashboards or multiple tabs where you've got, maybe, I've got my high level KPIs on the first page. The second tab is my sales KPIs, then my operations, then my HR, then my financials. So, it's OK to design it that way.
And then, certainly you want to use summarized information. There's no reason to get hundreds of millions of rows of data into a dashboard. It's really designed for aggregated information, and then as you decide to go into the detail, that's where you start getting into the transaction level information. Where you've got already existing reports or underlying decision support mechanisms to support the dashboards. So, those are really the best practices around designing and being successful with implementation. Thank you for your time, and if you have any questions, you can go to SearchDataManagement.com or www.BITadvisors.com.