Technology and Organizations

Today’s guest author is Tracy Allison Altman, who writes the blog, Evidence Soup, about evidence-based management: What it is, what it isn’t, and how to make it happen. Tracy is the founder of Ugly Research, a technology company that has created a presentation format called Tiny Soapbox.  Tracy is the lead-off blogger for an ad hoc series of posts to run here focused on the emergence of true, functional transparency in organizations.  Please follow her blog for a focused view (and some crazy counter examples) of evidence-based management.

Many people inside organizations are now participating more actively: They’re exchanging ideas, making collaborative decisions, and adding transparency to functions that are crucial to innovation and high performance. They’ve got the right corporate culture, processes, and technologies to achieve what I call strong connectedness. Good stuff.

But participation and connectedness aren’t enough. To perform at the highest level, people need to know what works (and what doesn’t). And to do that, they need evidence: To help them make decisions, evaluate new ideas, or design management programs. (What is evidence? It’s information that demonstrates the truth of an assertion. Evidence is what helps people make up their minds – data, statistics, research findings, business intelligence, expert knowledge, observations, facts, and anecdotes.)

Participation without evidence is pseudo-transparency. By connecting people, you get lots of transparency into what they are saying, but you don’t necessarily get transparency into what the evidence says. However, there are some things you can do to get more evidence into people’s hands.

For example. To illustrate the crucial role of evidence in the modern organization, consider a hypothetical enterprise software company that’s releasing a new product (we’ll call them MaxSoft). During the planning stages, product managers gather evidence in the form of market research and customer feedback, and use it to determine which features will go into the new release.

• If MaxSoft is a participative organization that emphasizes the importance of hard evidence, then various stakeholders (marketing, R&D, sales, user groups) will be able to contribute evidence during this phase, and comment on the evidence contributed by others. Conversations and collaborations won’t simply offer opinions, and won’t contain opaque, undocumented references such as “the research proves…” or “we know that…” or “studies have shown….”
• And if the company has an especially sophisticated approach to evidence, participants will be expected to look at research findings, etc. from sources external to MaxSoft, rather than relying only on evidence created internally. For example, evidence about human capital and organizational development, such as how to hire, motivate, and compensate the best people. (This might come from trade publications, academic journals, and industry conferences.)
• Because this is new product development, the MaxSoft product managers must allow room for creativity, risk-taking, inference, and experimentation. The evidence is there to guide them, not to make decisions for them.

After the product launch, more hard evidence becomes available. Sales figures and other performance data provide tangible evidence to assess acceptance of the new product. Business intelligence and analytics are sophisticated ways to pinpoint the most promising customer segments, and will help MaxSoft improve its performance.

There’s qualitative evidence, too. Sales people will begin repeating customer feedback, whether positive or negative. This is where transparency and discipline are crucial. Rather than encouraging ad hoc or free-form participation, MaxSoft can instead gather evidence methodically, making it more representative and therefore more useful to decision-makers. For instance: Use an input form asking specific questions (this works with customers, too). So instead of a sales guy saying “our customers hate feature x-y-z,” MaxSoft will be able to know how many customers really said that. A meaningful enterprise feedback program, combined with social CRM, will generate valuable evidence. To assess customer satisfaction, the team can also use text analytics and sentiment analysis: This qualitative information will supplement sales figures, survey results, and other quantitative evidence.

Later on, when course corrections are needed, people can look for evidence to help them adjust the product positioning, software development, or sales strategy. Again, looking outside MaxSoft – not just inside – will be essential.

So how do you make this happen? Evidence without connectedness is only pseudo-evidence: It does no good if people aren’t aware of it. Here are some suggestions for connecting people to the evidence:

• Set expectations that people will look for, and follow, relevant evidence.
• Develop guidelines for participating with evidence. Identify what types are typically appropriate in various situations, so people look for the right evidence at the right time.
• Use presentation formats and technologies that help people find, interpret, and contribute evidence more easily, regardless of which ‘silo’ they work in. (This is the focus of my work at Ugly Research.)

People + Connectedness + Evidence = Transparent Participation. Without evidence, people can participate in conversations about what really is working, or is likely to work, for their organization. They can come up with theories, make forecasts, estimate risks, and generate new ideas. But eventually, they’ll need some evidence to prove all that up.

Comments or questions? I’m tracy AT evidencesoup DOT com and @EvidenceSoup on Twitter.

Friday night Santa Clara University’s California Program for Entrepreneurship (CAPE) had the pleasure of hosting serial entrepreneur, Derek Dukes, co-founder and CEO of Dipity, and early (1996-2005!) employee at Yahoo!.  Great discussion moderated by Dan Aguiar, but two specific things stuck with me: (1) How Derek’s experience has led him to see value in Lean Startup, (2) How Lean Startup fits with the bottom-up, but evidence-based approaches to innovation and management that I think (hope) are gaining broad traction.

My still-early understanding of Lean Startup begins with Eric Ries’ blog,

Ries writes (excerpt below, my comments in brackets):

“What are the characteristics of a lean startup? One that is powered by three drivers, each of which is a part of a major trend:

1. The use of platforms enabled by open source and free software…. [The cost of starting a company has come down as much infrastructure and access/distribution is now available for free]
2. The application of agile development methodologies which dramatically reduce waste and unlock creativity in product development…. [See Slide 15: Split A-B testing. Continuous Deployment. Just-in-time Architecture and infrastructure. Use defects to drive infrastructure investments. This is evidence-based management for startups.  It also sounds a lot like the innovation approach Intuit is taking: As an innovation moves through Intuit’s Brainstorm process, innovators are asked to test hypotheses step-by-step.  This is in contrast to an older approach that seemed to ask innovators to “eat the elephant” in one sitting by starting with almost a full business plan.]
3. Ferocious customer-centric rapid iteration…. [Very hypothesis focused.  From Four Steps to the Epiphany: (p. 26): Customer Discovery Step-by-Step: State your hypotheses. Test “Problem Hypothesis.  Test “Product Hypothesis.” Verify.]

My belief is that these lean startups will achieve dramatically lower development costs, faster time to market, and higher quality products in the years to come. Whether they also lead to dramatically higher returns for investors is a question I’m looking forward to studying.”

Steve Blank’s work on customer development is foundational to the ideas of lean startup. He notes, “startups fail from a lack of customers, not product development failure” and “The goal for release 1 is the minimum feature set for earlyvangelists [sic].” The critical point being that you need feedback from your customers to generate key insights.

I also found value in Ries’ Top 5 Myths About the Lean Startup
and his talk for Stanford’s eCorner, Evangelizing for the Lean Startup

Derek said he’s trying out the lean startup approach in a project using his “free” innovation time. He mentioned that they worked for over a year before publicaly showing Dipity.  Knowing what he knows now, both from his own experience and from following the development of the lean startup ideas, he wouldn’t work that long without going to users/customers.

My intuition is that the practice of lean startup is another example of systems savvy.  Derek Dukes, Eric Ries, and Steve Blank see the value in taking a step back and assessing the overall system. They understand the technology (either the developing one or the one they’ll use in their production). They understand the organizational practice (and how past product development practices have been the death of products and companies). They understand the people/customers and how we need to understand by getting our, and our customers’, hands dirty and learning from the result (see my prior post on learning about systems savvy).  

Successful practitioners of the lean startup model will be constantly looking for how to effectively — and efficiently — weave together technology tools, organizational practice, with a deep understanding of the customer.

Thank you, Derek, for giving me this homework.  Readers, anything else I should add to my studies? Please leave a reply below.

Researchers generally like to tell us how things are. This is an important goal, but how does the process change if the real question is how can we make things better? Yes, we need to know how electricity works and what causes cancer, but the questions get more interesting for me when we, as researchers, move on to: How can I make electricity work better? …and certainly, how can we stop cancer? Generally you need to know the basic how question before you can dig into the more interesting change/stop/improvement question — but I try and always be pointed toward that change question from the beginning. I’m going to be watching for this change focus as I attend day 2 of the 9th International Conference on Mobile Business and the 9th Global Mobility Roundtable (ICMB/GMR 2010).

Yesterday I presented work Lynne Cooper (JPL) and Tad Milbourn (Intuit) and I are doing to track the outcome of adding mobile access to Intuit’s Brainstorm innovation platform. Our data shows the basics of “how things are” now that Inuit has and uses the Brainstorm tool to support their innovation activities. Short answer: Intuit has new innovative projects hitting the market with numbers and pace that are amazing. But that’s not our full question. We want to know what happens when you enable mobile instant/always on access to the Brainstorm tool. Does the process get even better? ….drum roll…. why does it get better if it does, and what other changes can leverage the outcomes even further? (We only have the baseline data now, stay tuned for what happens when the mobile feature is available.)

Monitoring alone does not help us improve our organization. Monitoring just lets us know that evolution had a particular outcome. Knowing the demographics of who does and doesn’t use the mapping features provided by their mobile phone isn’t the end of the question, and it shouldn’t even be a question to consider on its own. We must go further. The most interesting presentations yesterday provided the why — why did one demographic use a set of tools and other not…. and (though I didn’t see a paper covering this yet) how can people be helped to make good decisions about the features that will or will not provide them with value? Might a technology feature provide them value in some settings or if accompanied by a related organizational change? How can we help people have good systems savvy around these use decisions?

This is my second post focused on Providence Regional Medical Center of Everett, Washington and how I see their innovations as evidence of organization-wide systems savvy. That is, their efforts show that they understand both technical and organizational opportunities, and that they have the ability to weave them together into innovations — in this case, life saving innovations. I read about Providence in a BusinessWeek article and am thankful to Kim Williams, the Chief Nursing Officer of Providence and Judy Espedal, a Cardiac Critical Care staff nurse, and Dr. James Brevig, Director of Cardiac Surgery, for taking the time to tell me their story. Their examples are valuable because they give hope for healthcare innovation, they show us an exceptional process over a long span of years, and they help us see technology tools beyond computers and email. Here I will look at their approach as an overall practice of systems savvy, in Part 1, I provided a more detailed view of their “single stay” innovation.

Providence Medical Tower

Why am I certain that the Cardiac Surgery team has systems savvy? Because they haven’t made just one transformation — but several — each drawing on both technology and organizational practice to provide improvements in patient care. It is also important to understand that their innovations have been triggered by different forms of observation. In Part 1, it was the nursing staff’s observation of problems associated with how patients transferred to different units during the course of their care. In today’s examples: Blood conservation (less transfused blood is better in many cases) and skilled nursing education – the changes were triggered by statistical analysis suggesting room for improvement. Statistical techniques are themselves technology tools to apply in your organizational setting.

Blood conservation: In 2004 Dr. Brevig began the blood conservation program based on published research. He pushed for changes in surgical technique to reduce blood loss, changes in bypass machine settings to reduce the use of transfused blood, added a blood conservation coordinator to the staff, and worked with the critical care nurses on how to provide blood on outcomes, rather than routine. Follow-on analyses of patient outcomes verify the program’s success. From 2003 to 2007 the transfusion rate has decreased from 43 to 18% and the hospital stay time has been reduced by a half-day.

Skilled nursing education: In 2007 the cardiac surgery unit saw a bump to 12.2% of patients being readmitted to the hospital within 30 days. They used their analysis tools to dig into the issues and found that many of these patients were from skilled nursing facilities. Kim Williams says the team looked to what they could do to help these facilities take on patients after surgery. Providence developed an education program where a cardiac surgeon/critical care nurse team visit the skilled nursing facilities to help educate the staff on how to take care of cardiac surgery patients — teaching them about the common problems associated with readmissions — at no charge to the facility. In 2009 the readmission rate was down to 8.1%.

Once could be a fluke. Twice could be coincidence. Three times (single stay, blood conservation, skilled nursing education) is demonstrated skill with systems savvy. I would stand by this assessment even if there have been failures in the mix (though I’m not aware of any), given that their methods include long-term tracking and adjustments based on data.

This data appears to flow freely via collaboration, attention to communication, and long-term commitment. I asked Dr. Brevig about the distinguishing characteristics of the group – what, in his opinion, enables this team to be distinctively different in terms of their approach? He replied that the collaborative nature of the hospital pushes for buy-in and input across all members of the unit. My interpretation is that when true collaboration occurs in such a complex organization, technology and organization practice opportunities will both end up in the mix. We form teams to get diverse input, and sometimes, as in this case, it actually works.

Judy Espedal also emphasized the time and focus needed, noting that transformations take years and that you have to apply yourself from start to finish. Commitment is more likely given the Providence environment: With collaboration and communication, commitment is more likely.

From buy-in to showing that the innovation works, instituting the new practice, and on-going evaluation… these are conscious, explicit applications of systems savvy. No one functional area is making a decision. No one layer of the organization is making a decision. No one technology or practice stands above the rest. Providence takes a long-term perspective working with their full system of opportunities.

Summary of Providence’s systems savvy:

1. Ability to use different methods for identifying opportunities (not using just one lens).

2. Application of both technology tools and organizational practice to address the opportunity (the basics of systems savvy).

3. Long-term focus and tracked outcomes to suggest further adjustment (savvy, wisdom, is more than a short-term activity).

The doctors, nurses, and staff of Providence Regional Medical Center have a clear ability to see room for improvement and to find ways to do something about it. They draw on physical and analytic technology tools and make adjustments in their organizational practice given the variety of skills available, and they increase skills when needed. They do not try to fix a problem with a single “silver bullet.” The team uses an integrated approach where technology and practice support one another to reach specific goals.

I thank Providence for taking the time to share their story. This is health care reform.

People with systems savvy understand technical and organizational opportunities, and have the ability to weave them together to do exceptional things. I read about Providence Regional Medical Center of Everett, Washington in a BusinessWeek article on innovation in healthcare and it was immediately clear to me that the staff at Providence practice systems savvy on a broad scale. Kim Williams, the Chief Nursing Officer was kind enough to walk me through their process. Judy Espedal, a Cardiac Critical Care staff nurse, then gave me the history, and Dr. Jim Brevig, Director of Cardiac Surgery, the context. Their examples are valuable because they give hope for healthcare innovation, they show us an exceptional process over a long span of years, and they help us see technology tools beyond computers and email. This is a two-part story: In this post, I’ll describe one innovation in detail; in the next post I will look at their approach as an overall practice of systems savvy.

Providence Regional had been an early participant in “fast-tracking” routine cardiac surgical patients (e.g., patients with scheduled by-pass surgeries). Patients in the fast-track program would spend the night following surgery in the Cardiac Critical Care Unit and change rooms next day to the “step-down” unit to begin monitored recovery and rehabilitation. This quick transition to rehabilitation supports faster recovery and shorter hospital stays. Twelve years ago this was an innovative and successful program, but as the hospital grew and the patient load increased, they began to see problems:

Kim Williams: [We're] fairly proactive about how we manage patients and look at processes. We noticed that after the night shift nurse helped the patient up and the day shift nurse helped them to the room’s recliner in preparation for transfer, sometimes a bed [in the step-down unit] wasn’t immediately available.

Judy Espedal: I was noticing that patients were staying in the recliner for 4-6 hours waiting for their bed. By the time they did get to their room they were exhausted. The therapists assigned to them were gone for the day and patients weren’t receiving respiratory, physical, or occupational therapy until the next day thus missing out on 24 hrs of care. We are Critical Care nurses, not physical therapists. We didn’t have the practices built into our routines to get them moving [important for faster recovery]. We didn’t have the tools to walk them — they were hooked to everything.

[Around 2002] I had a hallway conversation with Jim [Dr. Brevig]… “they are missing a whole day.” By the time they get their new room, get in bed, and the new nurse assesses, it’s 4pm. They are missing a whole day of walking and other therapy.

Dr. Brevig’s response: “What can we do about this, Judy?”

Judy Espedal: So that’s when we started. We formed a task force to brainstorm ways to get around this. “Is there some way we can bring the care to the patient?”

That’s the origin of their Cardiac Surgery Single Stay Unit (CSSU). In a single stay unit the patient stays in place while the care and equipment come to them. The nurses, patients, and families have a single location that is changed to meet the patient’s needs. Kim Williams notes that they have portable X-rays, smart pumps (for medication), and telemetry units for each patient. As the patients improve the staff moves the equipment out to make the room look like the patient is getting closer to home. Judy Espedal describes it as “family centered care.”

They didn’t implement this approach on a whim, but based on data, and more data. Before asking for a pilot program Judy Espedal, Dr. Brevig, and representatives from hospital administration, the partner step-down unit, and respiratory therapy visited two hospitals that had implemented single stay units (though not for cardiac surgery). The team brought back impressions and outcome data from the other hospitals. Following the presentation of this material, Providence gave them two of the unit’s beds and permission to pilot the single stay approach for three or four months. They brought in portable monitoring equipment to let them extend their capabilities and added training for the nurses around how to switch from critical to rehabilitative care as the patient’s needs change.

Note that they had to weave together technology and practice: A single stay approach needs technology that is portable and staff with a broad range of skills to manage the different stages of care. While you can renovate to support the single stay (Impacting Patient Outcomes Through Design, pdf, Katherine Kay Brown), the Providence team found ways to work with what they had. Kim Williams believes single stay can work in any room with a toilet (yes, room architecture is a technology!) — meaning most hospital rooms could be single stay rooms if the other technology and practice adjustments are viable.

Judy Espedal says patients and families are now working with the same critical care nurses from admission to discharge. When a nurse can focus on the same two patients over days it allows for better care. “We can think of everything. We can get in an extra walk… sit with family… time to look at whole picture.” Dr. Brevig notes that information transfer across the team is also improved given their twelve-hour shifts: The increased rapport given contact time with patients and family means that subtle, tacit knowledge transfers more accurately — and only one “hand-off” takes place per day, rather than the two that would be required if the shifts were eight hours. Less opportunity for critical information to be lost.

The result of Providence’s ability to restructure based on technologies at hand, training to increase the breadth of nursing skills, and adjustments to patient care: Success. They decreased the time patients spend in the hospital by approximately a half day, have 99 to 100% satisfaction ratings from patients and family and numerous awards for the unit, and receive numerous requests to tell their story to other medical providers. Is this evidence of systems savvy or just a lucky outcome? Given they’ve made a variety of similar transformations over the years — I believe the staff of Providence Regional Medical Center are demonstrating full-on, team-level systems savvy. In Part 2 I’ll describe two more of their successes and summarize what I see as foundations we can all use.