On July 8 2024, Hurricane Beryl made landfall near the Houston metropolitan area. Beryl was a category 1 hurricane, the lowest rating out of 5 on the Saffir-Simson scale used to measure hurricane strength. For Houston’s electrical grid, which is managed mainly by the utility company CenterPoint Energy, the impact was anything but small.
Over a dozen people died. An estimated 2.5 million homes were left without power. In Houston’s humid subtropical climate, this meant being exposed to a heat index of 101°F (38°C) for days. Suddenly, millions of people in the United States’ fourth-largest city shared this top task: use CenterPoint Energy’s digital experiences to understand when their power would return. Unfortunately, CenterPoint Energy was not up to the task.
Top Tasks and Crisis Situations
Top tasks are activities that users must be able to do with a product. If users cannot do these things, then the experience has failed.
Apps and websites can be designed to enable many user activities. Identifying a short list (around 10 or so) of top tasks serves as a guiding light for the organization. What absolutely must users be able to do? How do you allocate resources to ensure those top tasks are likely to be successful in terms of design, engineering, or business investment? (Stakeholder interviews and user surveys are good techniques for determining this top-task list.)
Under normal conditions, the top tasks for a utility company might include signing up for an account or paying a utility bill. However, any organization that provides essential services, such as electricity, must always consider what their users must do when that service cannot be provided, particularly in times of crisis.
Being without power is not just about missing Netflix. Consider this hurricane situation and imagine the prospect of dangerous heat with unrelenting humidity. Your phone connection works sporadically at best, and your battery is dwindling. All stores within miles of you are closed. You are then confronted with a variety of challenges and decisions, such as:
- Will you have to throw away all your costly refrigerated and frozen food before it rots?
- How many restless, hot nights can you endure sleeping in your bathtub to stay cool?
- How much more heat can your elderly dog withstand as you watch them lethargically lay on the floor?
- Should you rescue your grandpa, who lives in the affected area and relies on an oxygen concentrator to breathe comfortably?
- When do you decide to flee your unpowered home because your baby is inconsolably crying in discomfort? Where will you go?
Faced with such stressful, uncertain situations, the top tasks normally lurking near the bottom of that list suddenly become crucial:
- Does CenterPoint know I’m without power? Should I inform it?
- What is being done to fix my area’s power situation?
- If I go somewhere else, will that destination have power?
- When will the power return?
From this author’s personal experiences, here were some UX problems with how CenterPoint supported these top tasks during Hurricane Beryl.
Problem #1: No Outage Map
In May 2024, a different severe storm caused widespread power outages in Houston. Due to performance issues with its previous implementation (the surge in visitor traffic suggesting how important this task is to users), CenterPoint removed its outage map from its website with the intent to replace it with a new one by the end of July 2024.
This was bad timing for CenterPoint. The decision to remove this capability during hurricane season without an immediate replacement rendered these top tasks heavily unsupported. This absence left users scrambling for workarounds, such as using the ordering app from the Whataburger hamburger chain as an indirect indication of areas with and without power.
Problem #2: Findability Problems with Outage Map
Realizing the lack of an outage map was a problem, CenterPoint deployed its replacement outage map. But finding the map on the website wasn’t easy at first, and it required scrolling past metrics and news posts to locate the link.
Clicking that button didn’t immediately take you to the “restoration” map. Users first had to scroll through a significant amount of text detailing CenterPoint’s repair process. While well-intentioned (a neighbor might get their power back before you do because of localized line damage near your home), this content was difficult to read and verbose. All of it interfered with the top task of quickly assessing your personal power situation.
But that’s not all. CenterPoint illustrated its repair strategy across several maps. Users had to choose the correct map to view, increasing cognitive load. They then could click on the map and begin interacting with it.
Problem #3: Outage Map Lacking Detail
After the users finally arrived at the map, they were greeted with a heat map (pun intended?) of the affected areas, which was nearly all of Houston. Users needed to interact with a button control to reveal a field where they could enter their address. When they found their home, the information was sparse, jargon-filled, and provided little decision-making support. Compounding the problem were reports that the map data was inaccurate and outdated.
Problem #4: Vague and Intermittent Alerts
Finally, CenterPoint’s experience frequently encouraged users to sign up for its Power Alert Service, which relayed text-message alerts regarding updates to outages and repairs. Using the text-message channel is a clever tactic, as users can read texts without internet access.
In the past, CenterPoint used to provide specific, reassuring details about the status of an outage. However, with hurricane Beryl, long periods passed without alerts from CenterPoint, until suddenly the power returned. For all the detailed talk about repair-process stages on the website, communication regarding these activities was absent in these alerts.
A Counterexample: Duke Energy
Other energy companies (who also operate in areas affected by hurricanes) inform their users about power outages and repair efforts more effectively. For example, Duke Energy (who won our Intranet Design Annual in 2011) has a well-designed outage map that provides users with centralized, actionable details written in simple language with clear, user-focused priorities. It’s unknown if Duke Energy could handle the widespread simultaneous impact of a hurricane like Beryl, but the experience seems better-designed.
Users Do Not Care if Their Top Tasks Are Difficult to Support
Energy companies cannot stop hurricanes and the damage they inflict (putting aside the preventative work they should be doing to mitigate damage in the first place). Heroic repair crews need resources and time to inspect and make repairs, so restoration estimates will always be approximate and never as quick as a user would prefer.
However, in coastal regions of the United States, hurricanes are predictable and forecastable events. Their threat to electrical grids is well understood, and losing power can quickly escalate beyond mere discomfort and become a matter of life and death.
CenterPoint’s user experience suggests a lack of internal preparedness or priorities. Removing the outage tracker with no replacement ready and the poor execution of its substitute suggest that CenterPoint leadership believes quality outage information is a nice-to-have task for its users. Or, maybe, engaging in service design and equipping repair and support teams with easy-to-use reporting workflows so that more information can be relayed to users are perceived as too much work.
Users seldom bother checking outage information during normal times. But when the crisis hits, that top task surges to number one on the list, and there are no excuses that will mollify anguished and confused people.
Conclusion
Companies, especially those that provide vital public services, must prioritize the top tasks of their customers, even when those top tasks are infrequently performed or inconvenient to support from a business perspective. Crises are defining moments for companies. Pass the test, and your users will endorse, defend, and praise you. Fail, and be cursed for a generation.
All companies exist to create value for their customers. When a company neglects its users’ top tasks, it has failed its fundamental reason for existence. Slowly, laboriously — yet inexorably — that company will be forced to pivot and serve its users, or it will perish.