Start by defining the outcomes. Work backwards from how you want the various stake- holders to describe the product after its been in production for several years.
Thinking long-term is vital as it could be years before you get to change course again. In addition to defining a strategy that will win in the future theres also likely to be a more immediate need, and that is to close the gap with the competition. A good road- map will cater for both.
With the benefit of hindsight, UIs built over the last couple of decades failed for a few specific reasons and its important to take onboard those lessons; i) Existing UIs were visualizations or projections of the systems functionality whereas they should have been a reflection of the users workflow, ii) UIs didnt cater for the future they solved for a point in time and so fast became unfit for purpose and iii) some skills required to build a great UI differ from the backend skillset.
In 5 years time, if you were to interview your clients about their experiences using the UI, what answers would you want to get ? Here are some examples:
Two sides of the same coin. The ultimate test for any new product will be the test of time which will depend on how easy it is to change and adapt but also how much it adds or takes away from overall profitability.
Having a solution thats easy to learn and be productive on will help in adding new de- velopers and make less experienced developers productive. Similarly, testability, de- ployability and supportability will be key given they can count for up to 66% of software costs.
Although the web stack seems quite stable new technologies will continue to emerge. Any new product design needs to be as modular as possible to enable simpler swap outs for better tech as it comes along.
Applies to both of the above but is also worthy of its own section. Polishing visual com- ponents can be extremely costly and inconsistent look-and-feel will kill the user expe- rience. A clear reuse strategy is required to keep costs down and agility up. Deciding on the right level of granularity and what is common is vital. Too coarse-grain will limit reusability but too granular will be hard to maintain.
Most front office users need a system to aggregate data from many sources, to be an- alyzed and acted upon. Historically this has been a predominantly manual activity and different asset classes are in different stages of migration to more digitized, excep- tion-based ways of working.
A new system needs to cater across the client spectrum, from users who are slow to adopt new approaches to users that want to be on the bleeding edge. Its pretty much a given that over the next 5 to 10 years, cost and regulatory pressures will push auto- mation as far as it can go and what gets built now will need to handle it.
A good question to ask is why do we even need to do this exercise? Ignoring budget restrictions, why didnt the UI evolve along with the backend functionality?
Generally, its because the UI has always been an afterthought, a lower priority than the backend. Historically UI engineering was not recognized as being technically challeng- ing and typical skillsets did not address the artistic and creative elements.
The need to maintain one user experience across multiple products requires org changes. Conways Law is a fact of life and although its possible to maintain more than one consistent experience across multiple teams, it will create more work and risk to do it that way.
Maintaining a consistent UX will come with an administrative cost. A functional, em- powered design council will be required to own UX integrity. One element of this will be deciding whether a requirement for a new visual component really represents some thing new or not. Clear taxonomy and design guidelines are required here to maintain the house view on whats common and whats special
Big Tech and the Internet have completely changed the landscape and expectations on what a UI can do and today consumer technology users have highly usable, works-of-art in their pockets. Keeping up requires much more than the traditional soft- ware engineering skillset and specificialized creative resources need to be added to the team.
Maximum configurability : Any and all visual formatting or style should be accessible by the user, from macro ele- ments like workspace, to micro elements like font type or row height. Users should be able to maintain multiple different configurations and switch easily between them.
End-client viewable : Client facilitation businesses should be able to configure and permission tear-off views on the fly and make them accessible to their end-clients, making interactions more seamless.
One UI with one way of working : For products with a broad set of user personas, irrespective of function, there should be a consistent experience, thats familiar and logical to the user. Like keyboard au- to-complete, the UI should be able to complete the users sentences.
MS EXCEL like : Excel is the most successful piece of business software ever. Copying and improving on some of the more powerful parts is a winning strategy. Its ability to enable non- technical users to create powerful tools and workflows on the desktop paved the way for todays low-code products.
For the grid components, imitating EXCEL features, like being able to create powerful derived data columns using formulas and multi-dimensional pivot tables, are a must.
Monetizing data : EXCEL is also great at enabling data monetization but struggles with real-time data. A new UI needs to make sure that any available data, should be accessible and usable anywhere and remain responsive, irrespective of the data volume. Examples are :
User oriented means workflow oriented : A modern system needs to be designed from the user down and the users workflow involves other apps on the desktop:
Serving the End-user that can code : End-user skillsets have seen a steady shift towards technology over the last decade. The rigidity of most desktop trading systems, combined with the need for rapid inno- vation, have forced able front-office staff to overuse EXCEL and other UDAs (User Defined Applications). The desktop tools of the future will need to provide a similar sandbox but with the right levels of control and governance (source code control, auditability, resilience etc.)
Decision orientated Going Beyond the grid : The grid or blotter has been the staple data visualization for at least 25 years. The blotter is data orientated and it drives a do-it-yourself workflow: here is the data, go find information and then act on your decisions. Modern UIs will enhance productivity by being decision orientated and exception based. The user can define rules, the system can operate autonomously and notify when it doesnt know what to do. Notifications would be highly actionable Here are decisions you could make, here is the data that supports the decision along with the ability to act.
Power search Going beyond the mouse and the menu : Mouse control in general is much less precise than the keyboard and finding func- tionality by traversing menus is hard, even when you know what you are looking for. All search should be possible from an intelligent text-based search function. Wheth- er you are looking for features to use, looking for trades that match certain criteria or setting up an exception-based workflow rule, all this can be driven much more effi- ciently from search.
Leveraging working patterns with machine learning : Most peoples working patterns are repetitive but also unique. User clicks, behaviors, searches should be stored, and machine learning algorithms employed to predict what a user is likely to want to do next and then render useful information and func- tions on the screen in anticipation.
Finer grain visual components: To enable much of the above the UI needs to be quite dynamic and able to group together different visual components. This means the traditional coarse grain com- ponents need to be carefully broken up into more fundamental building blocks that can be joined together in different ways. For instance, The order ticket needs to be 1 configuration of the 20 different visual elements.
At Velox our mission is to provide an application platform to our clients that maximizes their ability to innovate rapidly. Would be great to hear your opinions on factors that you think are missing.