Clean drinking water for Oksibil in Papua

When water is bad for your health

Health blogs, health coaches and health authorities recommend us to drink two liters of water per day. Still 50% of the world population does not have access to clean drinking water [1] and annually close to 1 million people die from water, sanitation and hygiene-related diseases. The lack of basic water and sanitation is estimated to cost $260 billion globally every year. [2]

When clean drinking water is not affordable

In West Papua, due to the high poverty, access to clean water is very low. In remote areas, like Oksibil in the Bintang mountain range, importing clean drinking water is not affordable, as its population of around 4000 people can only be reached by air. From 1970, five years after the area was opened to outsiders, to 1995, the dutch missionary Kees van Dijk worked in the area. He learned one of the six local languages, managed to earn the love and trust of the locals and started to represent them. This continued after he was asked to be based in the capital, Jakarta, in 1995. Still he regularly travels the almost 4000 km to Obsikil to help the locals.

Oksibil is situated in the Pegunungan Bintang mountains in West Papua at 1,307 m elevation and is only reachable by air – photo: Neneng Husein

Water, a gift from heaven

Fr. Kees is now trying to give what the people in Oksibil need the most: clean water. The country’s inspection agency Sucofindo (‘PT Superintending Company of Indonesia’) has declared the water in Oksibil unsafe for drinking. Geological samples in 2016 indicated that even at 90 m deep, only clay and dirt is found. Drilling this deep or even further is economically not feasible. As Oksibil on average has 400-500 mm monthly rainfall, collecting or ‘catching’ rainwater was the next option. A first Rainwater Harvesting (RWH) system with water storage in the ground was built, but this wasn’t resistant to earthquakes. Since 2017 three new earthquake-proof RWH systems with elevated water tanks have been implemented, but they are far from complete. And to serve the regional population minimally, two more will need to be added. After these RWH systems have been completed, the project can be expanded to other areas in West Papua.

Local ownership

To create local ownership and avoid operational dependency on sponsors, Fr. Kees, for building and maintaining the RWH System, insists on the “gotong royong” spirit of local community involvement. The locals help gather materials such as sands and gravel. Other materials like cement, pipes, filters, and storage tanks, have to be air-flown from the capital of West Papua, Jayapura, and especially from Jakarta. Due to the transportation, a sack of cement is 20 times more expensive in Oksibil than in Jakarta.

Water storage tanks, earthquake-proof elevated above the ground – image: Water for Papua

Helping the people of Oksibil

To build one RWH system, it takes 46 sacks of cement at a cost of US$ 4,000, pipes worth US$ 4,100 and a filter with a price of US$ 1,900. This sums up to US$ 10,000, excluding the cost of large capacity water storage tanks. To give the people of Oksibil clean drinking water, five RWH systems are needed. Dear blogreaders, tell your friends about this great project; all promotion and financial support are highly appreciated.

Contact and transfer information

Donations can be made by bank transfer only to: English-Speaking Catholic Mission, Neptunstrasse 6, 8032 Zürich, Switzerland; IBAN: CH 47 0020 6206 3632 1001 B, BIC: UBSWCHZH80A. Please mention as purpose “Donation Water for Papua / Fr. Kees van Dijk OFM“.

or to: Account Name: Van Dijk Cornelis G M; Bank name: Bank Mandiri; Bank Address: Kramat Raya, Jakarta; Account no.: 1230004105690; Country: Indonesia; Swift Code : BMRIIDJA

More information about the project can be obtained at www.facebook.com/waterforpapua and water.for.papua@gmail.com.

Oksibil is located at West Papua province, the far eastern part of Indonesia at its eastern border, south east of the Philippines and close to Australia. Indonesia from west to east measures 5,100 kilometres.

[1] https://www.solidarites.org/en/taking-aid-further/combatting-waterborne-diseases/

[2] https://cleanwaterfund.charity.org/

[3] https://www.wra.org.au/1609wpdcwater

Can Artificial Intelligence reduce the bias in Marketing and Market Intelligence?

In June 2019, Jake Silberg & James Manyika of the McKinsey Global Institute (MGI) published the essay ‘Tackling bias in artificial intelligence (and in humans)’ [1]. In the below article, Philip van den Berg shares his experience with this phenomenon in Marketing and Market Intelligence. He shares some thoughts about reducing relative bias and the state of ‘lack of bias’ or ‘absolute fairness’, including conventional ways on how to reduce bias and conclusions from the MGI article on how to apply AI to do so.

The Bias Dilemma for Marketing and Market Intelligence

An important dilemma for Marketing and Market Intelligence practices is often to identify, quantify and communicate bias, while maintaining credibility and business justification. Bias is defined as ‘the action of supporting or opposing a particular person or thing in an unfair way’ by ‘allowing personal opinions to influence your judgment’ [2] and includes ‘prejudice’, ‘statistically unexpected deviation’ and ‘systematic error’ [3]. The consequence is that market data, market insights and market segmentation, as well as marketing plans, marketing content and marketing actions remain debatable or even questionable. This occurs especially when business results are under pressure and marketing impact is below expectation.

I have seen senior management using a mix of three approaches for decision making and communication: data, stories and intuition. The first is often dominant: data driven managers use numbers to align people and to reduce bias. The phrase ‘data don’t lie’ is used regularly, but is this true? Silberg & Manyika show that not only data interpretation can be biased but also data itself is often obtained from a non-representative sample, with a subjective methodology.

In a more ‘siloed’ organization or partnerships, departments don’t trust the ‘fairness’ of each other and declare their own data source and insight as the best. The market intelligence analyst defends his research, the marketer or agency his competencies and expertise and the sales person his experience and customer relations.

Increasing fairness bias by transparency, omni-data and feedback

What a person does not know, he tends not to trust. A first step to create confidence and thereby to increase fairness is transparency. This starts by questioning: a) the data, b) the algorithms and analytics that turn data into intelligence and c) the interpretation or insights, in order to understand the bias. Here is important to document the findings and communicate them to the stakeholders that use the data, the intelligence and the insights. Most of the time, being open about bias and data quality limitations creates more trust, than just stating the are ‘great’ or ‘sufficient’. Transparency also encourages stakeholders to bring suggestions how to improve quality and to start co-owning the topic of improving fairness.

A second way to reduce bias is an omni-data approach, by efficiently extracting value from multiple data sources. With every source added, more data quality checks can be built in and insights become richer, deeper and better. Stakeholders who demand using another source to take away their remaining distrust, can in the in most cases be satisfied.

A third part which is often missing, is the thorough post-cycle or post-event feedback loop. It allows stakeholders to review, to what extent data and insight assumptions were biased and to agree with them, on where to improve and to take joint action.

Bias transparency, an omni-data approach and feedback loops lead to a better understanding of and more cooperation on how to increase fairness. This is not only valid for Market Intelligence but also for Marketing activities, from the market insight, the market segmentation, and the persona definition, to the marketing plan with the messaging, the marketing mix and the metrics.

To make the organisation bias-aware and capable of reducing it, a data-driven strategy and a culture of openness on data quality are essential. For this, leadership has to understand the value of fair data, to map where the organization is and should go and to start a transition project with a midterm horizon.

Reducing bias by experimentation

Advantages of starting a strategy and culture shift are, that they may take too long – the market, competition and customers don’t wait – and that they don’t state well, what fairness is. Silberg and Manyika conclude, that this last topic is so complex, that ‘crafting a single, universal definition of fairness or a metric to measure it will probably never be possible’. Instead they see different metrics and standards to be used, which each depend on the use case and circumstances.

Reducing bias however, means one still needs sort of an understanding of fairness and how to improve it. I see experimentation as a quick way, to determine how relatively biased for example a marketing campaign is. Testing and trying out different small scale scenarios in parallel on persona definitions, messaging and marketing actions, will provide useful insights and learning. The scenario with the best business result is likely to be the least biased one.

Reducing bias with Artificial Intelligence

Still, even the best scenario could still be biased and far from the point of ‘ultimate’ fairness. In seeking to identify this point and reduce bias, human behaviour and judgement have clear limitations. This raises the question, to what extent Artificial Intelligence, which has the promise to overcome human limitations, can help.

Silbert and Manyika see it as a challenge, that the underlying data are often the main source of the bias, rather than the algorithm itself. This is because the algorithms are often trained on data that contains human bias. The authors observe three main approaches to increase fairness in AI models, but conclude technical progress is still in its early stage. The first is data pre-processing for accuracy and independency reasons. The second is post-processing to transform AI model predictions to less bias. The third is including fairness constraints on the optimization process or using so called adversaries to reduce bias from for example stereotyping. Also adding more data points, innovative training techniques, like transfer learning and explainability techniques [4], can help.

Moving forward with Artificial and Human Intelligence

While clear definitions and the above approaches can certainly reduce bias, they cannot rule out fairness restrictions in the data collection or in the social context into which an AI system is deployed. Therefore the Silbert and Manyika state that ‘human judgment is still needed to ensure AI supported decision making is fair.’ This means that an adjustable mix of human judgement and AI judgment is needed. To find the best balance, in order to maximize fairness and minimize bias from AI, they recommend ‘six potential ways forward for AI practitioners and business and policy leaders’:

  1. Be aware of the contexts in which AI can help correct for bias as well as where there is a high risk
  2. Establish processes and practices to test for and mitigate bias in AI systems.
  3. Engage in fact-based conversations about potential biases in human decisions.
  4. Fully explore how humans and machines can work best together.
  5. Invest more in bias research, make more data available for research (while respecting privacy) and adopt a multidisciplinary approach.
  6. Invest more in diversifying the AI field itself.

Summary

The availability of almost ‘endless’ amounts of customer and business data, as well as the fast growing capabilities of Artificial Intelligence-powered data analytics, have brought Market Intelligence and Marketing into a new era. Companies were never more dependent on data as well as data analytics, and thereby on data bias and data fairness. These topics have become strategic and require a paradigm shift in the way organisations deal with them, with deep consequences for their strategy and culture.

This calls for the need to define the state of ‘ultimate’ fairness and to quantify the bias gap in both Market Intelligence and Marketing. This can be partially obtained by transparency, omni-data, feedback and experimentation, but these approaches have their limitations. While AI-powered data collection, analytics and enrichment solutions are still in an early stage, they add substantial value in reducing bias. As AI-generated data and insights also use biased data and biased algorithms, a flexible mix of human judgement and AI judgement is required. Although defining the ‘biassless’ or ‘ultimately fair’ state might still be difficult, this approach is an important step towards it.

The business value of AI will continue to increase in the near future. This will strengthen the competitiveness and the business results of companies and organizations. Therefore it is of strategic importance, that their C-suites embrace ‘Data Bias and Fairness’ as a strategic theme and start utilizing the ‘six potential ways forward’ of Silbert and Manyika.

[1] https://www.mckinsey.com/featured-insights/artificial-intelligence/tackling-bias-in-artificial-intelligence-and-in-humans
[2] Cambridge Dictionary
[3] Merriam-Webster
[4] While the high-performance and accuracy of Artificial Intelligence, that is Deep learning and Machine Learning algorithms, are generally valued, the models are often applied in a black box manner. This makes it difficult for researchers and data scientists to fully understand how the algorithms work, to understand how to assess the bias and define the point of ‘absolute’ fairness and to communicate the reason of the outcomes to stakeholders or customers. ‘By providing an explanation for how the model made a decision, explainability techniques seek to provide transparency directly targeted to human users, often with the goal of improving user trust.’ They consist of ‘ Local explainability techniques’ that ‘ explain individual predictions, which makes them more relevant for providing transparency for end users.’ and of ‘Global explainability techniques’ that ‘refer to techniques that attempt to explain the model as a whole.’ [5]
[5] Several authors; Explainable Machine Learning in Deployment, 13 September 2018; https://arxiv.org/pdf/1909.06342.pdf

Marketing and Market Intelligence: what is it all about?

The definitions of what Marketing and Market Intelligence include and what not, show a great variation and are sometimes not that clear. In the below article, Philip van den Berg shares his thoughts about this topic, where the two overlap each other, and for both provides a definition, that he thinks works best.

A great variety of Marketing definitions

Definitions of Marketing range from narrow ones, limiting it to planning and executing promotional activities, to broader ones, that include applying the complete marketing mix, exploring potential customer needs and defining the go-to-market strategy. Some discussions on the definition of what Marketing covers even want to include Sales into it. Originally Marketing was a company-centric term that tried to order market opportunities, customers and partners around the own organization. For a selection of definitions, see the list below.

In the last decade the maturation of e-commerce has shifted marketing towards Customer Experience, Loyalty and Advocacy. Consequently, attempts have started, to more successfully organize the company and its partners around the customer, in order to interact at all points of the Customer Journey. Direct customer-vendor touch-points concern not only Marketing but also Sales and Services. Indirect touch-points include partners, regulators, customer advocates, and other external and internal stakeholders. To ensure consistency and effectiveness in Customer Satisfaction and serving stakeholders, Marketing needs to cover all these groups within the own organization and in its ecosystem.

A comprehensive definition of Marketing

With the above in mind, I have come to this definition of what Marketing is:

Marketing is the discipline that, across the different functions of the organization and its ecosystem that have direct or indirect customer touch-points, aims to deliver exceptional customer experience and customer value, that attract and retain customers and create customer advocates by:

a) ensuring full understanding of the market potential, the customer needs, and competition
b) delivering the strategy and the plan of how to profitably realize that potential and fulfill those needs
c) executing the strategy and plan in an agile way, while reporting progress and final results

How about Market Intelligence?

Market Intelligence is the discipline that gathers, organizes, and analyses external data and delivers insights with the aim of supporting strategic or tactical decisions. It is an umbrella term that covers markets, competition, and customers. Intelligence of Markets, Customer Intelligence, and Competitive Intelligence use data to create insights. Therefore the term Market Insights would be better. For the best insights, external Market Intelligence should be combined with internal Business Intelligence.

Dax Sorrenti defines Customer Data as ‘the raw material of information about customers.’, Customer Intelligence as ‘the holistic and flexible understanding of customers that comes from gathering, contextualizing and analyzing data.’ and Customer Insights as ‘the deep understanding of customers that comes from gathering, analyzing and synthesizing customer intelligence. Insight goes beyond the “who”, “what”, “when” and “where” to tell us “why” customers behave as they do, guiding better business decisions and delivering results.'[1]

The number of definitions on Market Intelligence is much smaller than those on Marketing (see some definitions below), but all agree that its scope is quite wide. It includes gaining knowledge and insights on markets and market players in all areas. Market players include competing vendors, customers, partners, suppliers, government and regulators. Market Research, Marketing Intelligence, and customer feedback or reviews may all be seen as part of Market Intelligence. Therefore there is a big overlap in the areas covered by Marketing and by Market Intelligence.

A definition of Market Intelligence

These observations about Market Intelligence have led me to define Market Intelligence as follows:

Market Intelligence is the discipline that, across the different functions of the organization, aims to deliver meaningful insights for strategic, operational and tactical decision making, that allow delivering exceptional customer experience and customers value, outperform competition and bring structural business value and profitability.

Let me know if you have another view, have additions or simply agree!


Marketing Definitions

Philip Kotler: Marketing is the science and art of exploring, creating, and delivering value to satisfy the needs of a target market at a profit. Marketing identifies unfulfilled needs and desires. It defines, measures and quantifies the size of the identified market and the profit potential. It pinpoints which segments the company is capable of serving best and it designs and promotes the appropriate products and services .. .. the most important concepts of marketing .. are: segmentation, targeting, positioning, needs, wants, demand, offerings, brands, value and satisfaction, exchange, transactions, relationships and networks, marketing channels, supply chain, competition, the marketing environment, and marketing programs. These terms make up the working vocabulary of the marketing professional. Marketing’s key processes are: (1) opportunity identification, (2) new product development, (3) customer attraction, (4) customer retention and loyalty building, and (5) order fulfillment. A company that handles all of these processes well will normally enjoy success. But when a company fails at any one of these processes, it will not survive. [2]

Matt Blumberg: Marketing when done well is (a) the strategy of the business – its value proposition, go to market strategy, and brand positioning and image to the world. … Marketing in the twenty-first century must be (c) largely, but not entirely, measurable and accountable around driving business goals. Marketing when done brilliantly is driven by (a) includes a small, disciplined subset of (b), and is steeped in a culture of (c). [3]

Hubspot: Marketing is the process of getting people interested in your company’s product or service. This happens through market research, analysis, and understanding your ideal customer’s interests. Marketing pertains to all aspects of a business, including product development, distribution methods, sales, and advertising. [4]

AMA (American Marketing Association): Marketing is the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that have value for customers, clients, partners, and society at large. [5]


Market Intelligence definitions

Elli Mirman: Good market intelligence requires getting a complete view of a market, from marketing campaigns to product details to hiring activities. This includes: News: the latest announcements and mentions in the news, Team Intelligence: including job openings, employee reviews, and key leadership changes, Product Intelligence: product and pricing details, whether from product pages, help sites, app store updates, or other sources on or off a company’s website, Discussions: unfiltered discussions and feedback from customers and prospects on solutions they’ve tried, Marketing Intelligence: including content marketing, social media campaigns, and promotions across channels. [6]

KPMG: Market intelligence is the process of gathering, organising, managing, digesting, and finally delivering information with the aim of supporting a decision. … The scope of what market intelligence encompasses will vary from company to company, for example covering strategy, marketing, technology or other areas. We take a broad view of this scope, believing that companies’ needs and cultures vary widely, and therefore different organisations may benefit from different approaches to market intelligence. … Market intelligence covers one, some, or all of these topics (non-exhaustive list): regulatory changes, news, companies/competitors, technology trends. Depending on which topics it encompasses, market intelligence is also sometimes called competitive intelligence or marketing intelligence. Generally speaking, these functions are alike in technique, but pursue different goals. [7]

Adi Bhat: Market intelligence is defined as the information or data that is derived by an organization from the market it operates in or wants to operate in, to help determine market segmentation, market penetration, market opportunity, and existing market metrics. Market intelligence is a vital aspect to understand the state of the market, as well as helps collect competitor intelligence which in turn aids towards becoming profitable. … Market intelligence gathers data externally providing you a holistic view of the entire market and not just your organization. However, incorporating market intelligence with business intelligence processes will enable a company to have a holistic view of the ongoing corporate performance in specific market conditions. .. Market intelligence is closely associated with market research and can be explained in three simple parts as follows: Competitor Intelligence, Product Intelligence, Market understanding. [8]


Notes

[1] https://www.visioncritical.com/blog/the-difference-between-customer-intelligence-data-and-insight

[2] https://kotlermarketing.com/phil_questions.shtml#answer3

[3] https://www.recoveryview.com/Topic/TabId/107/ArtMID/657/ArticleID/1366/What-Marketing-Is-and-Is-Not.aspx

[4] https://blog.hubspot.com/marketing/what-is-marketing

[5] https://www.ama.org/the-definition-of-marketing-what-is-marketing/

[6] https://market-intelligence.io/leadership/what-is-market-intelligence/

[7] https://home.kpmg/lu/en/home/services/advisory/management-consulting/business-effectiveness/market-intelligence.html

[8] https://www.questionpro.com/blog/market-intelligence/

Book review: Strategy-In-Action: Marrying Planning, People and Performance

Transition-to-Success Framework for strategists & implementers, management & employees

Why do many corporations but also SME’s, scale-ups and start-ups struggle and even fail? It’s not because the initial business idea wasn’t good or the mission or organizational setup wasn’t good. The older companies get, the more they can create an environment where its employees want to crawl into the comfort zone of what has been achieved. One becomes defensive of historical ideas and positions. In the meantime, the business environment changes at what seems to be an ever faster and more disruptive pace. Strategies or often just tactics are adjusted more and more reactively and the responsible managers or consultants move to the next challenge before they can be held accountable.

Thomas Zweifel offers a methodology that joins strategy with planning and binds planning, people and performance. The goal is to be always open for the future and for transformation and capable of implementing it. The Strategy-In-Action methodology empowers the people in the organization to be future-focused, to get the relevant intelligence, to give room to deviating view, maximize buy-in of all relevant stakeholders, to get so-called quick wins, which are key to involve the organization in the transformation and to identify early people that my block the transformation.

The book shows how to bridge the gap between strategy and actions, and as I have experienced at large, inward-focused companies, between tactics and metrics. The methodology of Thomas Zweifel is clear and logical and will bring success. Implementing this will not only transform the organization to become more agile and future aware but also helps to get stakeholders involved. Therefore his book is a recommendation for both strategists and implementers and both management and employees at any organization, from established corporations and governmental organizations to smaller and younger companies in growth pains or decline.

Strategy-In-Action: Marrying Planning, People and Performance (Global Leader Series), Dr. Thomas D. Zweifel & Edward J. Borey

Driving purchase behaviour and customer loyalty

I have always been fascinated by what drives purchase behaviour and customer loyalty. Whether it was at the local dairy shop and the mobile supermarket, that I as a little boy visited with my mum, or at today’s rapidly growing high tech brands that are disrupting the world. What value, what experience do they bring, to make customers buy, become engaging, become and become advocates.

Analysing the market, building the right strategy, AI driven predictive analytics to ensure success and executing on it. … Marketing for Customer & Business Value … Optimizing Marketing & Sales Performance ..

These are the topics that drive my curiosity, that I like to drive and that I want to share about on this marketing blog.