RBC recently got in on the augmented reality act itself, creating its own AR app. It uses the camera, compass, and GPS to understand where the user is, and what they are looking at. It then marries this information with location data for its branches, enabling people to see an indicator representing the closest branch, along with the distance that the user would need to travel to get there. All of this information is superimposed over the video image on your smartphone screen.

Sarah Rogers, senior manager for mobile strategies at RBC, said the bank is excited about how the app could help its clients. “In the original app, we launched a branch locator,” she explains. That used a simple map to show customers where the nearest branch was, but RBC decided that it might be difficult to use, especially if a customer was from out of town, or was standing on a street corner and trying to navigate a confusing intersection.

“Now, if you want to walk east, you might want to know where the branches are that are east of you,” she explains. “It lets you turn the phone in that direction and then see where they are.”

CommerceLab asked the other banks in the Canadian big five (BMO, TD, Scotia, and CIBC) whether they were using augmented reality. Those that responded (TD and Scotia) were aware of it, but not working on anything, they said. That would come as no surprise to Srinivas Krishna, cofounder of AWE Company, based in Toronto. His firm creates augmented reality applications for a variety of clients, primarily in the retail sector.

AWE targeted financial institutions early on for its services, recalls Krishna. “Who wouldn’t? This is Canada – they’re the biggest customers on the block,” he says. “But we didn’t go there. Banks have a high degree of security requirement around whatever they do, and so it seemed like maybe that’s something we could go to a little later.”

AWE focuses on using 3D augmented reality characters in its work, as a means of enhancing customer interactions and other experiences. One example is its “Time Tablet” technology, which superimposes 3D animations on the real world, to transport tablet users back in time and let them view scenes from history. They can also interact with the characters, who are aware of their presence.

Krishna would like to see this implemented in a financial services environment. “One place where this could be applicable is for virtual ATM machines,” he explains. “There was a mechanical revolution that happened, which is what the ATM was for. Now, there is a digital revolution happening. As services are becoming more digitized, there are ways to improve the customer experience.” Why should you need to go to an ATM at all, when you can interact with a 3-D character in your home, he asks?

Making cash withdrawals and deposits aside, perhaps there would be some merit in being able to interact with a virtual teller in whichever environment you are in, reducing the need to visit a branch at all. The attraction for banks is obvious, as they could reduce a key operating overhead.

Kevin Kee, Associate VP for Research In Social Sciences and Humanities at Brock University and the Canada Research Chair in Digital Humanities, sees even more opportunities for financial institutions when they get on board. He believes that banks are already working on more advanced augmented reality applications.

“Augmented reality could be used for personal financial management to display a user’s financial situation vs what it could or should be,” he says, adding that wearable technology will play a large part in bringing augmented reality into the mainstream.

Google Glass, the wearable display that puts computer-generated images in front of your eyes and also uses a camera to relay video and take pictures, will be a key vehicle for new applications, he suggests.

Perhaps in the future, wearables will eliminate the need to carry credit cards as the user will be able to scan an item’s UPC code and have the amount automatically withdrawn from/or credited to a linked account, Kee muses. “Likely there will be apps that allow Glass wearers to scan a cheque and have it automatically deposited in their account.”

Rogers hopes that this will be away to help financial services companies work their way into more intimate relationships with customers. For example, a mortgage is isn’t the sexiest part of buying a house, she says. “It’s about your home, and your hopes and dreams for that home. How can we help you have more relevant information?” One way could be overlaying information to add value to customers about what they are seeing.

RBC seems truly committed to crafting these new relationships. In January, it released its RBC Wallet, enabling customers to pay for RBC Interact debit or credit card purchases with their phones.

It will take a while for people to be comfortable letting organizations beam more intimate customer interactions directly before their eyes. There’s a danger of feeling as though your bank is truly ‘getting in your face’. Experts in user interfaces, marketing, and law will all have to hammer out the issues. But game changing shifts in customer experiences don’t happen overnight.

The University last year pledged to make sweeping changes to the processes that it used to commercialize technologies, following a series of 26 recommendations in a report by the provincial auditor. These included a review of the structure of the University’s research centres, better education on policies and processes, and a complete review of its commercialization strategies.

The report, which was requested by the University, followed the launch of a lawsuit, in which it sued two companies that it claimed had unjustly claimed exclusive rights to carbon capture technology developed at the institution. The rights were non-exclusive, the lawsuit said. One of the defendants, based in Regina, issued a counterclaim, accusing the University of trying to change the terms of its deal.

A lack of due diligence

Legal spats like these can come back to bite researchers and their institutions unless significant due diligence is applied to research commercialization agreements, warns Tamsin Miley, a BC-based barrister and solicitor who operates her own law firm Miley Law, specializing in intellectual property and research law.

“The whole process is made more difficult if the IP issues are not addressed at the outset,” she says. “It’s very hard to repair IP defects, if you’ve done the innovation without consciously calculating that you might want commercial results to come out of it.”

Acknowledging contributions

One aspect that many commercialization agreements overlook is who contributed to a technology and may claim ownership or revenues from it in the future.

Research is generally a team-based effort, and institutions must understand exactly who contributed to it when creating a commercialization agreement, warns Angus Livingstone, Managing Director of UBC’s University-Industry Liaison Office. “You can be walking into situations where you don’t have certainty of title,” Livingstone says.

Even when researchers think that they’ve covered all their bases, students who worked on a project can often be overlooked. Then, in the future, they may return to stake a claim after a technology makes it to market.

Understanding who worked on a project is only one part of the challenge, explains Miley. Researchers must also understand the affiliations of those on the team. One danger is that a team member’s funding comes from a government department, which could then become a stakeholder in the technology. Involving the government body can slow down the commercialization process, she warns.

Sector-specific considerations

All of this can be a particular problem when the IP is researcher-owned. On the other hand, some institutions operate on an entirely university-owned model, where the university retains ownership of the ideas, says Livingstone. This can speed up the commercialization process, making it easier to navigate the clear title that many commercial partners need to succeed, he suggests.

This can be particularly useful in certain industry sectors, such as computer software, where the market is very fast-moving, and being first to market involves a race.

On the other hand, a university-owned model can be useful in some sectors, such as biotechnology. Whereas software IP can be protected by copyright as much as patents, patenting is everything in biosciences, Livingstone says. Patents can be expensive, costing $15-$20,000 per patent. Researchers will often need a university’s financial assistance.

Publishing before patenting

The need for patent protection can also cause issues with funding and publishing models. ‘Publish or perish’ is a common phrase in academia, and researchers rely on frequent publication to help secure grant money for their projects. But once an invention has been publicly aired, the clock starts ticking on patentability; they must apply for the patent within a year of it coming into the public domain.

The patent application can also be long and arduous, and researchers must ensure that there is no existing prior art which could render their invention unmarketable. This task often requires an entirely different skill set.

Giving away too much

Dr. Michael Owen, Vice President of Research, Innovation and International at the University of Ontario Institute of Technology, offers one final warning: assuming that all of these other hurdles can be overcome, researchers must not give away too much when they license a technology. This includes the right to use it in their own labs.

“Researchers don’t necessarily recognize that they may be giving away all the rights to use the IP,” he says. “They must always retain the rights to use the outcomes of the research for their own research purposes.” This should also include the right to publish the results of the research, he says.

What’s clear is that Canada could do a lot better in its patent activity; the latest World Intellectual Property Organization (WIPO) statistics show the country ranking 12th in patent filing activity for 2012, behind the Netherlands, France, Germany, China, and of course, the USA. This covers patents applied for both inside and outside academic institutions; still, perhaps a little more innovation is needed?
photo credit: Horia Varlan via photopin cc

Spinning off a technology into a startup company can offer high returns, if researchers are willing to put in the time and effort to market it. But this approach can be harder for some technologies than others.

Choosing the right technology

“In order to attract investment, investors need to see an ROI through an exit strategy,” explains Peter Garrett, president and CEO of Innovate Calgary, the technology transfer and business incubation centre for the University of Calgary.

“Pharmaceutical startups can be challenging due to the long timelines to market and high cost of development,” Garrett continues. “Many energy spin-off companies can also have the challenge of high startup costs like those associated with field testing.”

Ownership models 

It isn’t just the type of intellectual property that determines the possibility of a spinoff. IP ownership structures may also be a barrier.

A six-university study by UC Davis professors Martin Kenney and Donald Patton suggested that inventor-owned research (as typified by the University of Waterloo’s model, for example) creates a far more conducive environment for spin-off operations. Conversely, while not necessarily preventing spin-off formation, an institution-owned IP model can put roadblocks in the way.

Building the right team

Assuming that researchers have the right intellectual property and business proposition to spin off a company, they must then think about skill sets. Even if the technology is world-beating, it won’t get far without the necessary business acumen, warns Paul Rivett, director of client services for VentureLab, a regional innovation centre in York Region, and a member of the Ontario Network of Entrepreneurs.

“Researchers often assume that they can make business decisions, without business experience, while they continue to pursue their academic careers,” says Rivett. “Researchers often overvalue technical expertise and IP, and undervalue market experience and business savvy.”

One way in which this can manifest itself is in a poorly-formed team, run by engineers rather than business experts. In some cases, says Joel Liederman, vice president at technology commercialization venture MaRS Innovation.

Often, the researchers may not even want to be a part of the spin-off company, knowing their own limitations and wanting to stay in the lab.

All of this can lead to a catch-22 situation, warns Liederman, who argues that it can be hard to hire the right team until a company has enough funding, but that the funding won’t always come until investors see the right team in place.

Venture funding

The context for funding isn’t good anyway. Canada is not a particularly fruitful place for raising investment capital, warns Liederman.

“There’s a dearth of investment capital in this country. We do provide seed funding, but we can’t go much beyond a couple of hundred thousand dollars,” he warns. Even then, the firm counts on funding partners like IRAP and NSERC to help it out.

“We have an exciting company that we helped to spin off from a university, and right now they’re in California in the YCombinator incubator,” he laments. “There’s a good chance they won’t come back.”

Venture capital investment in Canada still has not recovered from 2007, when they stood at $2.1bn, according to data from the Canadian Venture Capital Association. They have crept up since then, but still fall short, and haven’t shifted from $1.5bn for two years.

The amount of VC funding awarded to a Canadian company lags that of their American competitors. In 2012, startup Canadian firms got just 44% of the VC funding enjoyed by firms south of the border.

Don’t rely on the institution

One might argue that universities could step in, but as Rivett argues, this isn’t a guaranteed outcome. “Presumably the academic institution would take into account the risk of obtaining sufficient investment in determining if the spin-out company could be successful,” he says. “I would not expect the institution to step in, to fill a gap – it would not be a sound investment decision on their part.”

Tom Corr, president of Ontario Centres of Excellence, says that this is why it is important to get industry as involved as early as possible. “We can tell you from the start if you’re on the right track because we get the voice of industry involved when reviewing applications for funding,” he says.

Having advisors with some business smarts is just one strut of a broader strategy to starting a spinoff company, as is having a well-rounded team. But ideally, the researchers responsible for the technology itself will themselves have a track record in engaging industry.

The more seasoned the senior players, the more sympathetic the money men will be – and in any case, researchers trying to get technology off the lab and into the market might want to set their sights on lower-level angel funding, rather than shooting for millions of dollars out of the gate.

Previously on CommerceLab: The sure-fire ways to succeed (or fail) at technology commercialization

Photo Credit: star5112 via Compfight cc

The problems range from the financial to the legal. Joel Liederman, vice-president of MaRS Innovation, recalls one professor who came to his organization with a technology that could have been a potential market hit. Sixteen universities and research institutes collaborate within Liederman’s organization to negotiate technology deals between research groups and commercial entities. He should have been the perfect conduit. “We would have people that publish the paper, come to us 11 months later and then say ‘here’s my invention disclosure’. We would then say ‘you realise that we only have one month to secure patent protection?’”

Solutions looking for a problem

One of the most prevalent issues around commercialization is a misunderstanding of business goals when forming commercial partnerships, warns Tom Corr, President of the Ontario Centres of Excellence, which helps to commercialize projects from publicly-funded research organizations.

“Often when researchers are developing IP to commercialize, they do it in a void without any help from industry,” says Corr. “Sometimes they are disappointed to learn when they try to commercialize it that, in fact, there may not be a market for it or the market may be so small that nobody’s going to invest in it to move it forward.”

In short, innovations within a university may end up being a solution in search of a problem. A failure to understand this can leave good research sitting on the shelf, explains Corr, which can be frustrating for researchers who were convinced that their projects should be the Next Big Thing.

If they’re lucky, a commercial partner may be willing to try and retool their technology for the market. But this may involve them trying to tweak or ‘pivot’ the technology in a new direction, warns Liederman.

MaRS Innovation tries to streamline this process using the Framework Funding Agreement, in partnership with commercial companies. It works in two ways, he says. The organization can present a portfolio of projects that companies can pick through, to try and find the most appropriate ones for them. Commercial partners can also stipulate their strategic priorities, which then prompts MaRS Innovation to field proposals from research teams in those areas.

When companies steer research

Liederman doesn’t deny that this financial “carrot” can sometimes steer the direction of research projects looking for funding.

“The lifeblood of the researchers in our member organizations are the dollars to fund their research. Those research dollars – the unfettered ones from NSERC where you could do whatever the hell you wanted – those are becoming harder to get,” he says.

But technology databases are overrated, argues Kevin Dunn, Director of Industry Liaison and Innovation at Dalhousie University, who says that he has operated them for a while but never commercialized a research project using that approach. He calls commercialization a “full contact sport.”

“Someone has to engage and discuss, sometimes even translate the science into the business applicability,” warns Dunn. “Someone needs to put some business speak in normally. Some researchers can do that, but it’s a rare researcher I’ve seen who can.”

This is both a question of education, and of personality. MaRS sometimes educates researchers in the commercialization process, teaching them to understand business needs. But sometimes, researchers will try to spend some time working for a company that is commercializing their own technology, and find themselves driven back into academia, points out Liederman. Others enjoy the process so much that they never look back.

In some cases, researchers can flit between the two worlds, working for several companies during the course of commercializing their research, taking a leave of absence to run companies or assist there. Those individuals, who understand the strategic priorities both in the lab and the boardroom, can truly thrive.

Focusing too narrowly on an industry sector

A Technology Transfer Office is one option for researchers who don’t understand the business needs of a potential partner. These university departments can grease the wheels and help bring businesses and researchers together, bridging any gaps in knowledge and terminology that may arise.

This engagement includes understanding how a technology can be repurposed for different sectors. A chemist may think about commercializing a product with a large chemical firm, but might not think about its applicability in aerospace and defence, for example.

“Their innovation may be of much higher value in an area that they may not have been exposed to,” Dunn suggests.

Problems letting go

If a partnership does grow, some scientists can have problems letting go, and adjusting to the realities of a business environment, warns Dunn. Those unrealistic expectations can go both ways.

“Sometimes the commercialization partner wants way too much control. They want to own the researcher and have complete control. Or the other side is that the inventor can’t appropriately disconnect from the technology. They want to become the CTO of the company, and that can be very limiting.”

Misaligned academic and business goals is just one potential hurdle for commercialization projects. In the second and third parts of this series on CommerceLab, we’ll explore how researchers struggle with intellectual property challenges, and the harsh, cold realities involved with spinning off their own companies.

The CFI has launched a search engine designed to help companies find research projects that they could use commercially.

Called the Research Facilities Navigator, the site lists 300 labs and facilities across Canada. Companies can search by topic (such as ‘engineering’, ‘environment’, and ‘health’), along with keywords. They can also narrow the focus down to a province, a city, or one of the 59 institutions included in the database.

The Navigator came out of a strategic roadmap developed last year by the CFI, a non-governmental organization that funds research facilities across the country. “One of the strategic directions is to promote strategic partnerships between institutions and users of research,” said Pierre Normand, VP of external relations at the CFI.

Labs will have their own fee structures for companies wishing to those their facilities, but there are also opportunities for deeper collaboration between researchers and commercial organisations.

The research landscape in Canada is shifting quickly, says Maria Lantin, director of the Intersection Digital Studio at Emily Carr University of Art + Design. She was among the first to register on the search engine, and hopes that it will be a useful tool as she pursues commercial partners for her facility. IDS’s Variable Frame Rate Technology was used for the Desolation of Smaug, the second film in The Hobbit trilogy. Her studio also collaborates on rapid prototyping and 3D printing with commercial partners.

Research cycles are speeding up thanks to tools such as social media, which are making technology conversations more visible, says Lantin. “Going from idea to product is faster, so companies have to respond with that same kind of speed,” she points out. “A lot of companies don’t have many resources to devote to research and development.”

Companies won’t be the only organisations using the site to find researchers. Eric Bosco, chief business development officer at Mitacs, says that he has a pilot team of business directors that will be testing the system in its work with clients. Mitacs specializes in connecting researchers with commercial partners.

“The more searches we can make and the more data points we have got, the better a match we can make,” says Bosco.

There have been other efforts to connect researchers with commercial organizations using technology. The Alliance for Commercialization of Canadian Technologies (ACCT) provides a members-only tool enabling companies to search for 110 technology offices in Canadian research institutions. ACCT spoke to the CFI in the early days of the Facilities Navigator project and had hoped to collaborate, but the CFI initiative was already too far advanced, says ACCT president Janet Scholz.

But building relationships goes beyond technology. Scholz says that connections between researchers and commercial organisations are made in two broad ways. One is on a scientist-to-scientist level. Many longer-term collaborations result from a chance meeting at a conference, or one scientist seeing another’s published work. The other way is via the marketing of the technology via an institution’s technology transfer office.

“I would always caution people to pay attention to the detail,” she continues. “Sometimes, scientists get very excited about doing something, and then go home to their organisations and find that they can’t.” IP transfer rules, freedom to operate and privacy rules can all get in the way, she says.

Academics with experience of commercial partnerships are already aware of such potential pitfalls. Dr. Jurgen Rehm of the Centre for Addiction and Mental Health highlights avoiding conflict of interest, clearly defining both sides’ objectives, and sufficient dialogue as three critical success factors in any relationship spanning the commercial and academic research worlds.

But it shouldn’t stop people from forging those partnerships, says Rehm, who runs a Mobile Research Lab that companies can use to assess employee health status. “We should not restrict ourselves to the ivory tower, but systematically explore how much we can change in the real world,” he says.

Being able to connect companies, institutions, and researchers so that these changes can begin is an important step toward the goal of commercializing Canadian research. It isn’t the only one, though. It paves the way for sophisticated relationships between two different worlds, that can be challenging to build and manage. Technology may help make the connection, but beyond that, the challenges are intrinsically human.

Rockwell, Director of the Canadian Institute for Research in Computing and the Arts at the University of Alberta, is co-leader of Digital Games for Learning and Training (DIGLT), a project that focuses on researching the mechanics of good game design.

DIGLT is part of the GRAND network of centres of excellence, a selection of research projects across Canada. It takes complex digital media concepts and try to create products of use to users across a variety of disciplines. Along with DIGLT co-leader Christina Conati at UBC, Rockwell’s part in that is to work out how games can be used to learn.

That doesn’t necessarily mean playing them. One of the best ways to gamify learning is to create environments where students model complex knowledge by designing a game, rather than simply playing one, Rockwell argues.

“We started with the hypothesis that designing a game is the best way to learn. Thus we are trying to understand how games get designed in serious contexts,” he says.

Rockwell has been developing two projects designed to help students design games, but says that it is hard to prove, and expensive to develop. “The one we are getting the most traction with is actually a card game because it is really easy for participants to design their own deck of cards,” he says.

The pair have also been developing an assessment framework of questions that people could ask themselves as they develop a game. But isn’t the concept of gamification already covered? Companies such as Bunchball are already selling products that translate gaming concepts into corporate environments, for example. They have ‘playbooks’ outlining the fundamentals of game mechanics.

But Rockwell says that they’re “light on research.” Gamification isn’t new – it’s been in the classroom for years, he points out. Any teacher who ever kept a score board for classroom behavior can testify to that.

“What we want to do is study when it works and for who and under what circumstances. Bunchball and others are selling products – hopefully we can help them sell better products based on real research.”

Location-based games are a promising area for the project, says Rockwell. DIGLT is using a variety of location-based tools, including SCVNGR, a location-based gaming platform that imposes a game layer on the real world (and which also had its own playbook). DIGLT students used this to develop Return of the Magic, a game that involved going into stores (video here).

Rockwell is also interested in how games can be used to teach more effective writing. “We have developed a prototype of a gamified writing environment, but are now adapting it to work with a particular writing course so we can actually try different types of game features,” he says. But this project highlights just how difficult that is to do.

The software would support hundreds of students learning how to write, but at the same time would have to support the writing instructor’s particular requirements. “Each instance of gamification tends to be customized to the situation,” Rockwell says.

Rather than commercializing the project’s own games and competing with industry, Rockwell and his colleagues are interested in developing Massive Open Online Courses (MOOCs), which can teach those destined for the private sector how to use that gamification knowledge to best effect. “Universities are too slow to do commercial research,” says Rockwell.

Perhaps in learning the fundamentals of good game design, students will be able to take Rockwell’s research and run with it, creating wealth in the Canadian economy. If he can succeed in that, he’ll consider it a solid win.

Earlier this month, for example, the credit card firm rolled out the latest version of Financial Soccer, a website using an online soccer game to teach people financial literacy, and then test them on it. Players successfully answer questions to pass the ball in the FIFA-themed contest. The Canadian edition features updated questions and is available in both English and French.

Money management and financial product information can be overwhelming, says Carla Hindman, director of financial education for Visa Canada. “So it’s about bringing it to people in a way that’s fun and engaging, and to start the conversation.”

Others are doing similar things. The Investor Education Fund (IEF) runs Cranial Cash Clash, a game that tests financial knowledge. Like Visa’s, there is a leaderboard so that people can share their scores, motivating them to play. And in September, it launched ‘Dollar Decisions’, a scenario game aimed at teens trying to save.

One of the benefits of gamification is that it produces lots of data, which organizations can use to help hone their messaging. “This is a research tool for us,” says Perry Quinton, marketing VP at the IEF. Cranial Cash Clash was produced using questions from the organization’s financial literacy survey.

“We know from that what areas people are weak in and need different information in,” she says. “We can aggregate it and use it to direct our resources.”

The banks are also playing with gamification to encourage financial literacy. ING Direct offers Lil’ Savers Moneyland, an online resource for children, featuring games to teach budgeting. Mark Nicholson, head of digital, creative and communications there, says that a separate piggy bank connects to the app digitally, and uses coin reader technology to track a child’s savings in the account.

For younger kids, a generation that is hyper-connected digitally, platforms like Lil’ Savers puts money and other aspects of financial matters in a context and format they understand and are comfortable with, he says.

RBC offers an iPad app, Learning Money with Leo, which teaches money concepts to 3-6 year-olds. And its Daily Advice Game on Facebook targets adults with video content, inviting them to submit the best answer.

“Using elements of gamification can help deepen engagement in educational content, help convey complex concepts in easy-to-grasp ways, and encourage ongoing participation and dialog in financial literacy content,” an RBC spokesperson said.

Perhaps. But gamification in financial literacy is still rudimentary, complains gamification consultant Gabe Zichermann, who organizes the GSummit gamification conference.

“The weakness is the idea that knowledge is a key differentiator,” he says, arguing that simply turning a school curriculum or survey into a game won’t cut it. “Sometimes, knowledge is very divorced from action.”

In this sense, financial literacy is similar to health care, he says; people often don’t make short-term sacrifices (such as saving, or giving up smoking) because the benefits are long-term, and the incentives aren’t high enough to offset short-term pleasures.

“So, some [financial institutions] have spent lots of money building financial literacy games that have no effect,” Zichermann says. Instead, they must put long-term resources behind behavioural change, Zichermann says.

That’s a far more sophisticated kind of product. It would be personalized, and would probably draw heavily on mobile apps and big data. But it would provide feedback and coaching that would affect user actions, he argues. It’s a world away from random questions about the definition of compound interest, and rewards badges.

Zichermann cites Nike Plus and Fitbit, tools that track your activities all the time, give you positive reinforcement around your fitness goals, and advising you on what’s needed next. “Someone needs to make the FitBit for finance.”

The company announced late last month that it would include previews of pictures and Vine videos directly in its users’ mobile and web timelines. That may have made advertisers happy, but not all users liked it.

“Have to say my Twitter feed is getting irritating with all the full-sized photos, etc.,” tweeted Mathew Ingram of GigaOm recently. “If I wanted a photo stream I would go to Instagram.”

How careful should start-ups be when making user interface changes, and how can they avoid irritating too many people?

“Change isn’t free,” says Paul Hibbitts, a user experience consultant in Vancouver who works with enterprise clients. When tech companies change their user interfaces, users have to invest effort—however minimal—in relearning something. There has to be a disproportionate benefit to make it worthwhile, he warns.

Finding those benefits is key. “Hopefully, everything every business does is in response to (user feedback),” says Shaun Illingworth, managing director of Toronto-based user interface design consultancy Akendi. “When you change an interface, just like when you change a product line, you’re doing it because of some kind of data point that you’ve perceived, whether it’s good or bad, from the marketplace.”

But where does that data come from? Usability testing—where users are formally observed using the product—can be one source. Other options include analytics, where user data from real user sessions can be mined for insights.

This data can tell you how long it took to complete a particular task on the site, and how many people completed it at all, for example, says prairie-based Blaine Bertsch, a usability consultant who has taught UI design at the University of Edmonton, and who runs his own startup, Dryrun.com.

“You might find that people are using different features and functions more than you thought, so your priority list changes,” he says.

But in some cases, the benefit of a UI change may be to the company, rather than the user. Advertiser-friendly moves are a necessary part of a monetization platform for a firm which badly needs to prove itself.

“It’s a delicate balance,” says Jesse Spink, creative director at Vancouver-based Ayogo, which specializes in gamification software designed to improve patient compliance in the healthcare market. Spink is constantly trying to bridge the gap between the needs of users and commercial stakeholders in his products, and trying to serve both at once.

“Conflicts of interest start to arise when people are very engaged and used to the software, like Twitter, but then the business needs evolve,” he says. How much you consider user needs depends in part on the company’s business model, too. Twitter is free, after all. “If you provided this amazing product (like) Twitter, and it’s augmented in new ways (by adding images), then that’s the cost of gate admission,” argues Spink.

Where companies are persuaded to make interface changes, they can at least do it gently, say experts. Twitter flipped the switch on its content embedding without giving users an option. Contrast that with Google, which offers people the chance to flip between old and new interfaces for free services like Gmail, which it kept in a long beta period, for a limited time. 

“I like a preview period where we give users the option to choose when they want to take an early peek at what we’ve got cooking,” says Hibbitts, He adds that companies should build in a feedback mechanism, enabling users to easily communicate their experience when they’re trialing a new interface element. And explaining why changes have been made isn’t a bad idea either.

These pieces of advice are worth remembering, especially when interfaces get so bogged down with changes over time that they have to be ripped out and replaced, to reflect a new information architecture or underlying business logic. As with customer issues, the key for tech startups is to think like a user—and to listen to what they’re telling you.