Headlines regularly announce the latest scientific breakthroughs which promise radical improvements in the fight against illness.

They are eye catching: DEMENTIA BUSTER New wonder drug hailed as a ‘game changer’ in battle against Alzheimer’s, or Miracle made in Britain! How the microscopic substance graphene can make sea water drinkable and even fight cancer, and from a few years ago Personal Genomes Will Spawn Made-to-Measure Drugs. But the medication that my doctor prescribes can be tens of years old and sometimes work and sometimes doesn’t. So why does it take so long for a scientific breakthrough to make its way into an effective medication ?

To get from the eureka moment in the laboratory to an effective medication is very complex. First there is the scientific discovery which establishes new facts , or explanation, that answers a question: what is the cause of cancer ?, why do cells divide ? and why is the blue print for our bodies wrapped up in something called DNA ? Once something new has been discovered it has to be turned into something that can be used. This can require developing new: skills, equipment, processes, and techniques. Finally, it has to be developed into something that people can easily use to cure, or manage, their illness.

Straddling all the stages from scientific breakthrough to medication are many regulations, intellectual property rights, industry standards, etc. some of which may require changing before the medication can be used. Then there is the tension between science and commerce where ones priorities is to discover new scientific facts and the other wants to take the breakthrough and make money from it ( the gap between the two groups is often called the ominous “valley of death” ). Many steps through layers of complexity add up to a long time and lots of money before the new medication is available for use.

To illustrate the complexities in taking a scientific breakthrough into a medicine it is interesting to trace the history of something sits on most cupboard shelves - aspirin. The story starts around 400 BC in Greece, when Hippocrates gave women willow leaf tea to relieve the pain of childbirth. It took until 1823 for the active ingredient to be extracted from willow and named salicin. Then in 1853 Salicylic acid was made from salicin by French scientists but it was found to irritate the gut. It took another 40 years until German scientists found a way to reduce its irritant properties. Then through the late 1890’s a process for synthesising aspirin was developed, clinical trials completed and aspirin was launched. The application of aspirin is still explored today, for example in the reduction of cancer risk. Today it is the best known and most widely used medicine in the world with an estimated 100 billion tablets taken every year. The history of aspirin shows how long it takes from the initial scientific breakthrough to being prescribed by a doctor, or bought of the shelf.

Digging behind the headlines shown at the beginning of this post their underlying status can be found. For the Dementia Buster claim the results are at the Phase III of clinical trials ( there are 4 phases of clinical trials) However, research into a cure for Altzhimers is notoriously difficult with success declining at each phase of clinical trials. Fingers crossed ! In the announcement of using Graphene to cure Cancer - university medical teams are working with it to produce minuscule drug delivery systems that can penetrate patients’ tumours before releasing cancer-killing medicines. However it is at the laboratory phase and a long way to go before clinical trials ! Using the Genome for personalised medicine there has been more progress. Specific genetic disorders have been been identified, for example most inherited cases of breast cancer are associated with two abnormal genes: BRCA1 (BReast CAncer gene one) and BRCA2 (BReast CAncer gene two). There is also progress with gene-targeted cancer drugs and in particular helping to identify targeted cancer therapies for a wide range of cancers. Also, companies are popping up that can analyse an individuals DNA for a relatively low cost e.g. 23andMe but linking the results to medical conditions is still work in progress. Personalised, or precision, medicine continues to be a promising area of research let’s hope that it gets lots more support.

Every scientific breakthrough should be celebrated - increasing our understanding of the world and ourselves can only be a good thing. But caution needs to be applied when the breakthrough is proclaimed through media headlines - a dose of reality needs to be applied and realistic estimates given when we can get access to the new medication. Also, it will prevent embarrassing discussion with my doctor when I start “But I have read about …”.

I use the web as a source for learning - see Using The Web To Learn - about a wide range of subjects: making sourdough bread; writing software to analyse a shopping basket; or understanding why football teams are no longer using the 4-4-2 formation. However, there is one area that needs to be improved: access to relevant information based on my level of knowledge about a subject.

Let me give you an example. I have reached an age when medical breakthroughs are becoming more important. Recently a breakthrough was announced in the fight against the terrible disease Alzheimer’s. I was curious about the science behind the breakthrough, in particular the analysis of the data to support the claim, and a quick search found a copy of the paper.

The paper was full of terminology that I didn’t understand. More searches helped to clarify the different proteins, scanning techniques, assessment of Alzheimer’s etc. that give an overview of the research. However, I was still interested in the statistical analysis which was used to support the claim that the test is 87% accurate. This is when the problems started. Further searches started to produce a mish-mash of information: poor quality videos, badly written overheads for a lecture, chapters from draft books, chapters from books no longer available, lecture notes, different mathematical notation etc. After many coffee drenched hours I eventually found the correct level of information that helped me to understand the breakthrough. However, instead of trawling through piles of confusing information it would have been better if my search brought back the information at a level that I could understand in enough time to drink one cup of coffee !

This is my challenge to the search engine companies - when searching on a particular subject I need the information to be at a level that I can understand. So search engine companies - sharpen up your algorithms, speed up your computers and lets see what you can really do !

On a good day I can answer two of the questions that come up during the BBC's programme University Challenge: the quiz where university students answer questions on a wide range of subjects. Using my smart phone I can answer more - it depends on how fast and accurate I can type ! Answering questions using the web is a powerful thing. But what if I want to learn more about a subject ?

I often come across a subject when reading about a hobby, developing a new skill or sparked by curiosity, where I need to find out more. But the subject can be like a brick wall built of strange words and unfamiliar concepts. I am stuck ! For me there is only one approach and that is self-learning, in other words, working through the subject by myself. The following is the way that I use the web to help me learn.

I start by identifying the people who are working on the subject and search for what they have written, for example blogs, papers, public presentations, books, videos on YouTube etc. I also search out any commentators, or critics, on the subject. This initial phase is like detective work - pulling together scraps of information about the subject and assembling them into some sort of overview.

Let me give an example. Tim Harford wrote an interesting article about our inability to predict the impact of new technology on how we get it wrong when imagining the effects of new technology on the future of everyday life. In the article he mentioned an interesting idea proposed by the economists Daron Acemogula and David Autor where we should view work in bite-size chunks - tasks rather jobs. Basically they think that “routine, codifiable tasks” can be automated while purely human skills such as problem solving and creativity can not. Searches produced papers by the two economists with titles such as: Skills, Tasks and Technologies: Implications for Employment and Earnings, Polanyi’s Paradox and the Shape of Employment Growth, and The “task approach” to labor markets: an overview ). I found Tim’s email address on the web and contacted him for advice on the best papers and he kindly suggested Why Are There Still So Many Jobs? The History and Future of Workplace Automation ).

The next stage is to immerse myself in all of the information that I have gathered which feels like drowning. But after surfacing for a few coffees I find that some of the new terminology makes sense and the subject starts to take shape.

Eventually I reach point where I feel confident enough to test my understanding of the new subject. This is done by working through some examples, or tutorials, and then using my new knowledge on a problem. This can be done more easily in some areas than others, for example using open source software I can test my understanding of new concepts in computer technology. For other subjects then I can only read about it e.g. surgery.

Continuing with my example, the subject falls under the heading economics. Therefore applying what I have learned I can draw any sensible conclusions about changes in business ? For example what is driving the increase in call centres ? Part of the job of a salesperson was to make regular calls to customers to try and drum up new business or chase repeat orders; this was usually when selling to a business and is still practiced today. However, call centres take the “task” of calling customers to retail customers and probably reduce the cost per call due to centralisation of the task. Not a rigorous conclusion but the start of a few interesting thoughts.

What I have described above is the first step. To gain a greater level of expertise in a subject can take many years: there is a debate raging on the web that suggests that around 10000
hours is required to gain a high level of expertise in a subject which is a long time but exploring a new subject is the fun of learning.

Next time I sit down to watch University Challenge I hope too have gained more knowledge from my self-learning so that I can answer more than my usual couple of questions - if not I can always access the web !

Ralph Bana, credited with many innovations in the travel and leisure industry, was quoted during an interview with the industry magazine Tour & Travel: "I just don't come in and say "ah-hah" ! The answer relates to clearly defining what the problem is, forcing yourself to write it down, researching the alternatives, picking one that will ring the bell and then going out and selling it". This is great advice that could be used by any industry, government, or organisation.

When a problem is written down on a piece of paper the sentences can be analysed in a way that fuzzy concepts, wild ideas, and general waffle can be removed. So how would this work ? Let’s take an example, the National Health Service ( NHS ) in the UK are often criticised about the speed of innovation. This is a complex problem and there maybe many underlying factors but it is often heard that there is a lack of money to help speed up innovation. One attempt at writing out the problem could be:

“The NHS is slow at innovation because it doesn’t have enough money.”

The first thing to observe about the statement is that it identifies a problem: “The NHS is slow at innovation” and a solution: “it doesn’t have enough money”. Digging deeper there are three concepts that are at best unclear: “slow”, “innovation” and “enough”. Starting with “slow” it is unclear how slow is slow - is it a perception or are there some hard facts ? ( a quick web search produces no hard facts about the speed of innovation in the NHS ). Next “innovation”; another search on the web shows that there are many definitions of innovation ( around 1,980,000,000 hits ! ) and therefore scope for many experts and management consultant to debate its definition for many years ahead. Finally “enough” - how much is “enough” ? - twice the current amount, three times the current amount ? Therefore we can conclude that the stated problem, and solution, is loaded with confusion and can only lead to further confusion, and result in wasting time and money.

A better example of analysing a problem when it is written down is:

“One of the highest priorities for the NHS is the reduction of obesity which is estimated to cost £6bn per year”.

The description of the problem is clearly defined and quantified. However, it could be criticised for being at too high a level - too abstract. It could be broken down into factors that are contributing to obesity, for example: education, economic, social context, life style, public health etc. which could be quantified, and then alternative solutions explored.

Ralph Bahna was always looking for competitive advantage, which had to be “neither short-term nor flimsily opportunistic”. The technique of writing the problem out and critically analysing its content is an effective way to clear out muddled thinking. With a clear problem definition then there is a greater chance of finding a successful solution.

A new digital business seems to hit the headlines everyday. The news report suggest that the entrepreneurs had a sudden flash of insight, and after a few coffee fuelled weeks of writing code, a product was launched and truck loads of money rolled in. Reality can be very different.

I recently came across a company called RentHop, an apartment listing service based in New York City, during a predictive modelling competition held on Kaggle. The aim of the competition was to predict the number of inquires a new apartment listing would receive based on the listing’s creation date and other features. To understand the context of the data supplied by the the competition, I gathered as much information on the company that I could find. Themes started to appear during my research that provided a fascinating insight into the hard work and determination that goes into the development of a very successful digital business - there are many lessons for the rest of us.

RentHop was started in 2009 by Lee Lin and Lawrence Zhou following bad experiences with brokers when trying to rent apartments. They initially thought that brokers were not adding any value and therefore the industry was ripe for a radical change - a disruption. After all if books can be bought, taxis ordered and restaurant tables booked with a few clicks why can’t a similar process be in place for renting an apartment ?

The first lesson to emerge from RentHop’s history is to understand the intricacies of the market for your idea - detail is everything ! To understand the detail it is important to experience the market that you are targeting both as a customer and as supplier. In the early days of RentHop they were challenged by the venture firm Y Combinator, who supported them during a summer incubator program, to spend time as brokers so that they could check if their idea was going to work. Lee and Lawrence gained their brokers licence, which required 75 hours of class, and then spent a few successful months as brokers in New York City. They learned the craft of a broker down to the nitty gritty, such as buying six-packs of Coronas for the Superintendents to get in with them so they could get a copy the master set of keys and therefore be the first to get into apartments. During their time as brokers they realised that the most important part of brokering is the relationship developed between the broker and renter. A really good broker makes the process a whole lot less painful for the renter and they realised that a great broker adds value. Therefore RentHop changed direction - pivoted - and worked on adding value to the broker’s activity by increasing the effectiveness of the leads. For example, at the heart of RentHop’s offering is the HopScore which rates the quality of an apartment listing. Listings with higher HopScore appear at the top of the search results where they’ll have more exposure to renters.

Nothing stands still is the next big lesson. Markets are constantly changing: new entrants, new technology, new legislation all provide a turbulent background where your are trying to hold onto your customers and find new ones. In the case of RentHop new entrants into rentals are AirBnb who are changing the rental landscape. Is AirBnB a short term let or are they moving apartments into the hotel business ? The development of mobile technology continues having an impact. It’s inherent value can be exploited, for example brokers arranging meetings with clients while on the move or updating the information about an apartment after a visit and therefore increasing its HopScore while on the move. Also, as RentHop continues to grow it can leverage its property data - its digital assets - and by combining with with public data ( e.g. New York City Rents By Subway Stop 2017 ) to provide a better service and therefore competitive advantage.

One advantage of a digital business is that it can grow very quickly compared to a bricks and mortar type business - easier to scale. However, another lesson is that scaling has to be sensitive to local markets. For example the HopScore had to be changed when covering other cities. In New York City the more people that see an apartment but show no interest, then it can be concluded that it is poor quality, and therefore the HopScore can be dropped. However in other cities there maybe bottom feeder investors looking at the apartment therefore the eleventh showing could be just as important as the first visit.

Behind the click-to-cash headlines for a digital business there are many years of hard work by dedicated entrepreneurs. Analysing the development of companies like RentHop provides valuable lessons for future entrepreneurs.

For more information on RentHop, its history, challenges and many more lessons, see:

You’ve Got The Power With Lee Lin, Founder of RentHop #137 - a fascinating interview which covers why and how RentHop started, the intricacies of rental markets, the impact of new technology and legislation. The post: Why Real Estate Brokers Exist in 2016 And Beyond analysis the future levels of automation of the brokers activities. Finally, Lee describes the similarities between brokers and chess players in the post: Hello from RentHop – and why brokers and chess players are similar