Be Data Solutions

Be Data Solutions is my latest venture which I have cofounded and run as CEO.

Be Data provides end-to-end expertise in data strategy, data engineering and data analytics. 

Our team has decades of collective experience working for some of the world’s most innovative organisations. 

If you need help or advice on any aspects of your companies data, whether that be strategy, infrastructure, architecture, engineering, analytics or data science, then drop me a message.

shorful@bedatasolutions.com
www.bedatasolutions.com

CEO/Co-Founder

Subatomic Analytics

I am the founder of Subatomic Analytics, an analytics and data science consultancy. 

I advise organisations on how to use data and analytics within their organisation. I have worked with large corporates to small start ups.

Also, through a partnership with Tribal Worldwide, Subatomic Analytics powered the Tribal Data Science Practice in London and Singapore between 2018 and 2023.

https://www.campaignlive.co.uk/article/tribal-worldwide-london-restructures-challenge-consultancies/1456169

https://www.ddb.asia/catchfire/tribal-worldwide-asia-launches-total-experience-tx-offering/

Founder and Managing Director

OutThink

Data & Behavioural Science Advisor

Headquartered in London, UK, OutThink transforms the way security teams engage with employees. Our award-winning cloud (SaaS) platform enables large, complex organisations to Identify, Understand, Measure and Manage human risk, in the context of cyber security.

As Chief Product and Data Officer I lead the product vision and set the data strategy.

OutThink is world's first human risk intelligence platform.
https://www.outthink.io

Chief Product & Data Officer

NELSON COLLEGE LONDON LIMITED

Non Executive Board Member

Tribal Worldwide London

Tribal has partnered with Subatomic Analytics (originally Stream Intelligence) to enhance its analytics and data science capabilities. I lead the Tribal Worldwide London Data Science Practice from a strategic and operational aspect, managing the P&L, resources and deliverables.

Tribal is the Total Experience Agency and Tribal Data Science helps organisations deliver value from the data they collect.

Chief Data Scientist & Practice Lead

Stream Intelligence

From being Managing Director of Stream Intelligence I have now moved away from the day to day mangement of the company to being an Advisor as the company evolves in its journey.

Advisor

Dr Shorful Islam is Co-founder & Managing Director at Stream Intelligence.

Stream Intelligence is a data and analytics consultancy that answers business challenges through the intelligent use of (big) data and advanced analytics.

We approach our client engagements strategically, by identifying the key business challenges and objectives, before diving into rigorous and robust data analysis.

Our people combine knowledge of statistics, psychology, economics and business strategy, and apply it to solve a wide range of business challenges including; targeted marketing, optimal pricing, growing product penetration and retaining customers.

Our belief is that businesses and organisations sit on a wealth of data that can uncover wider social and economic trends, which will prepare them for the future.

Co-Founder and Managing Director

University of Indonesia

Visiting Lecturer

Wunderman UK

I led the Data and Insights division at Wunderman UK

I managed a team of Database and Platform Operators, CRM, SAS, Digital and Social Analysts, Data Strategists and Planners, and Campaign Developers and Managers as well as Analysts and Campaign Managers offshore.

KBM Group is the global leader in knowledge based marketing solutions and is recognised year on year by Forrester as one of the top marketing service providers in the industry.

The consolidation of KBMG data practices into Wunderman provides the benefits of best-in-class data management and optimisation capabilities in an integrated, seamless and strategic fashion.

Allowing the Wunderman UK network to offer clients a broad, customisable spectrum of the data and marketing services that are critical to today’s digitally oriented marketing campaigns.

Managing Partner

Responsible for strategy, digital and social analytics planning as well as growing the UK business.

Director of Digital Analytics

DDB UK

DDB UK is part of DDB Worldwide Communications Group Inc . (www.ddb.com) the largest consolidated advertising and marketing services global network in the world, according to Advertising Age. In 2009 DDB was named Global Network of the Year by Campaign magazine. 

As Head of Data I look for data opportunities, ranging from helping clients understand their consumers better, spark ideas for briefs, develop models which help with the execution of campaigns, or power products.

Working with brands such as VW, Unilever, Kwik Fit, Virgin Media, and ExxonMobil requires an understanding of how different organisations collect and use data.

Head of Data

STV Commercial

Consultant to Commercial Director on Digital and Data Strategy

Digital and Data Consultant

Having worked across client side, agency side as well as public bodies in the area of data, analytics and insights, I have helped organisations use data to deliver results. I have worked as a researcher and analyst to being on the board of organisation, so understand the detail and bigger picture of what an organisation is hoping to achieve. 

I have always had to create and grow teams in all my roles, from growing and running a data and insights division for a large agency of over 100 people to creating my own data science team of around 35.

Currently, CEO of Be Data Solutions, Managing Director of Subatomic Analytics as well as advising a Cyber Security startup, on their product and AI capabilities and serving as the Chief Data Scientist at Tribal Worldwide London.

I enjoy helping companies use data and advanced analytical techniques to run a better organisation. I love working with analysts, strategists, creatives and techies, getting my hands dirty optimising analytical methodologies, brainstorming solutions, or sitting at the board with C-suite to help them determine how they can realise their vision and help them create a strategy to deliver it. 

I'm ultimately an academic who loves working in a commercial setting, driving results, change and innovation.

A great day at #BigDataLDN, interesting to see so many AI tools and providers.

Was great catching with people, Mujahid Rashid, Yusuf Raja and listening to some talks.

Dr Shorful Islam

CEO & Co-Founder | Data & AI Advisor to CEOs | Author of "Data Culture"

Another year at #BigDataLDN and great to see my book at the Carruthers and Jackson book stand.

Time probably to write another book

My team was blamed for a project's failure.

But it had nothing to do with their skills or their work ethic.

I once had a client who asked for a big report. They needed data from the last 12 months across all 40 of their international markets, and they wanted it within three days. It was a reasonable request.

The problem wasn't the work itself. My team could write the code to pull the data easily.

The problem was the tool we were forced to use—an old, slow internal database. When we tested it, we found that pulling the data for just a single month in a single market took six hours to run.

You can do the maths: 6 hours multiplied by 12 months, then multiplied by 40 markets, comes out to 120 days of non-stop work.

We explained this, but the client didn't fully grasp that the system was the bottleneck. They were disappointed when the three-day deadline passed, and it resulted in them feeling like the team had failed them.

This is a classic situation I've seen many times. 

The failure wasn't the people; it was the infrastructure they were given. A great team can't outperform the limitations of bad tools, but they are often the ones who take the blame.

P.S. Have you ever been in a situation where the tools you had to use made it impossible to meet expectations

In the world of data analytics, there is often a rush to the finish line, to deploy the latest machine learning models, to build complex predictive algorithms, and to uncover hidden insights.

While the ambition to use advanced analytics for business value is commendable, many organisations skip a crucial foundational step. This oversight leads to: 
- misguided efforts, 
- misaligned strategies, and 
- failure to generate meaningful return on investment.

Before a single line of code is written, you must first understand and map the customer journey. What you’ll get from this foundation:

1. You avoid solving the wrong problems.
- If you jump straight to advanced analytics without a clear understanding of your customers' behaviour, you often find yourself solving the wrong problems. 

You may build a churn prediction model, but fail to understand why customers are leaving.

2. You turn raw data into powerful strategic assets.
- Without context, data is just a collection of numbers. Advanced analytics, in this scenario, is akin to a surgeon operating without knowing the patient's medical history. 

You may be technically brilliant, but the outcome is likely to be suboptimal.

3. You identify the moments that truly matter.
- Journey mapping helps you pinpoint the critical interactions where customers are most likely to drop off, become frustrated, or become brand advocates.

These are the moments that truly matter. They are where your analytics efforts should be focused.

4. You create cross-functional alignment naturally.
- The exercise of mapping the customer journey forces different departments, marketing, sales, customer service, and product, to collaborate. 

It breaks down data silos and creates a shared understanding of the customer.

5. You can formulate informed hypotheses about behaviour.
- When you understand the journey, you can formulate informed hypotheses about customer behaviour. Instead of a fishing expedition with your data, you can now build targeted models to test specific theories.

My experience over two decades in the data and analytics space has taught me one simple truth: the most successful data projects start not with technology, but with a deep, human understanding of the problem. Every time. 

Before investing in expensive tools or hiring data scientists, invest in the fundamentals. Your data will then speak clearly, and your analytics will deliver lasting business value.

Customer Journey Mapping

Early in my career, I learned an acronym that has stayed with me: HiPPO.

It stands for the "Highest Paid Person's Opinion."

It describes a dynamic many know but few dare to name: when a senior leader's personal belief carries more weight than the data. This isn't just a minor frustration; it's the single biggest cultural blocker to becoming truly data-driven.

I've heard countless stories from data leaders and analysts:

Entire meetings were derailed because a senior stakeholder saw data as an "inconvenience" to their worldview.

HiPPOs advocating for strategies based on personal anecdotes—like a spouse's feedback—while dismissing concrete data that proves them wrong

Teams are forced to allocate precious time and resources to initiatives that the data clearly showed wouldn't work.

The most damaging part? It completely shuts down learning. The HiPPO isn't engaged in a dialogue. They issue opinions, and the organisation is expected to comply, creating a culture where analysis is theatre and facts are irrelevant.

This isn't about vilifying leaders. It's about recognising a critical cultural failure. With all the will in the world, one HiPPO can undermine an entire data transformation.

The shift to being data-driven isn't just about building models and dashboards. It's a shift in power—from opinion to evidence, from hierarchy to curiosity.

As leaders, we must check our own biases at the door. We must create environments where the best argument wins, not the highest-paid argument.

Have you encountered a HiPPO?

5 critical dimensions determine whether your data supports reliable decision-making or misleads you entirely.

Whether you're analysing customer behaviour, tracking sales, or optimising marketing campaigns, the quality of your data directly impacts the accuracy and reliability of your insights.

To minimise errors, build trust in your findings, and make confident decisions, you must consistently evaluate your data against these 5 pillars:

1. Accuracy: Does Your Data Reflect Reality?

A customer purchases an item for £9.99, but the transaction records show £-9.99. This negative value could represent a legitimate refund or a systematic error. Without validating against real-world records, you're building analysis on potentially false foundations.

2. Completeness: Do You Have All the Data You Need?

Missing every tenth transaction or having gaps in your website visitor data fundamentally skews results. Incomplete data undermines analysis validity regardless of how sophisticated your methods might be.

3. Consistency: Are Your Data Sources Aligned?

Your coupon redemption data updates every 48 hours, whilst transaction data updates every 12 hours. Yesterday's analysis shows current transaction data but lagged coupon data - leading to incomplete insights.

4. Timeliness: Is Your Data Available When You Need It?

Some retail stores upload transactional data daily; others take up to 14 days. If you're measuring marketing campaign impact immediately, waiting two weeks renders the insights practically useless.

5. Validity: Does Your Data Fit Expected Formats?

A £102,000 transaction at a grocery store warrants investigation. It's either genuine or indicative of deeper systematic issues affecting your entire dataset.

Achieving 100% perfection across all five dimensions is often impractical.  Focus on what's "good enough" for your specific use case and prioritise improvements with the greatest business impact.

Do you evaluate your data against these five critical dimensions?

P.S. Data quality isn't a one-time task—it's an ongoing commitment.

5 Critical Dimensions of Data Quality

3 reasons your 'humble heroes' leave. 

And it's rarely about the work itself!

As you grow your data team, you'll likely have a "humble hero". This analyst is the backbone of your team. They diligently work on every request, solve problems proactively, and produce robust, consistent work.

I call them humble heroes because they never seek recognition. They gain satisfaction from a job well done.

But that doesn't mean they don't need recognition.

I've spoken to these heroes about why they leave, and here are the three real reasons they give:

1. They don't agree with the data culture. 
2. They don't align with the vision for data.
3. They're not receiving recognition for their hard work.

Because they don't ask for praise, they are often overlooked. The first sign of their dissatisfaction is often their resignation.

Now, the solution seems obvious: recognise and reward them.
But the mistake most data leaders make is waiting for official approval before saying anything. In large organisations, promotions and budget approvals take time, so they stay silent, not wanting to promise what they can't deliver.

Meanwhile, their humble hero interprets this silence as indifference. They emotionally check out and resign.

The fix is simpler than you think.

Just take them aside after they produce or present great analysis and reassure them that their work is being appreciated—not just by you as the data leader, but by the wider organisation.

You don't need budget approval to give recognition. You need budget approval to give rewards

Don't let them emotionally check out while you're fighting for their promotion behind closed doors.

Have you ever lost a humble hero?

Most companies can't make the output of any ML or AI model actionable.

The failure isn't technical. 

Rather, it's a fundamental knowledge gap.

They skip developing a deep understanding of what their customers actually do. This understanding forms the foundation for sustainable business growth. Rather than spending a fortune on complex technical solutions, focus on the value hidden in your existing customer data.

At Be Data Solutions, we help companies unlock the insights already sitting in their customer data. When you truly understand your customers, three transformations happen:

1. Tailored Solutions 
When you deeply understand your customers, you can tailor your products or services to meet their specific needs. For one client, we helped them understand that a segment of customers only came to their department store for cosmetics. This enabled the merchandising team to curate cosmetic product offerings which resonated with that segment.

2. Effective Communication 
Knowing your customers enables precise communication using language, tone, and channels that resonate. We experimented with formal and informal tones for one client and found a significant uplift in click-through rates when using "Dear" instead of "Hi".

3. Enhanced Customer Service 
Understanding customers' pain points and purchase history transforms service delivery. Working with a customers service team, we empowered them with customer insights so that when a customer called, they knew what the customer had previously purchased and the likelihood of the call being a complaint or query. This enabled the team to respond appropriately and improve customer satisfaction.

Sustainable business growth isn't built on complex tech.

It's built on deep connections with the people you serve.

P.S. Are you focusing on deep customer understanding first?

Actionable ML and AI

Most organisations approach data literacy the wrong way.

They roll out company-wide training programmes, expecting everyone to become Power BI or Tableau experts overnight. I've seen this countless times - teams get sent on generic data literacy courses that cover the basics but miss the mark entirely.

This approach falls short because sending everyone on a general data literacy training course isn't always feasible or practical, as it may not capture the specific nuances of how data is used in your organisation. 

Each business has its own metrics, its own definitions, and its own way of interpreting data.

Solution? Create a Data Glossary.

A glossary of data terms should clearly describe each metric in business-friendly language. It should explain how metrics are calculated and interpreted, with examples if necessary. 

Think of it as your organisation's data dictionary - but written for humans, not machines.

Before diving into creating your glossary, keep two important points in mind:

1. Specificity: The definitions of metrics and KPIs will be unique to your organisation. Your "conversion rate" might be completely different from another company's.

2. Dynamic Nature: The glossary should be a living document, constantly updated by a selected group of people as needed. Metrics evolve, business priorities change, and your glossary should reflect that.

Don't overcomplicate the creation process.

If your organisation already has reports or dashboards, use them as a foundation for your glossary. 

Start there - identify the metrics people are already looking at and define them clearly.

"We've always done it this way" is the most expensive sentence in business.

But it has nothing to do with your employees.

It's a symptom of a culture that has lost its 'why'.

In my research for Data Culture, I came across a powerful story often called the ‘five monkeys’ experiment. While its scientific origins are debated, its lesson for business is undeniable.

Scientists place five monkeys in a room with a ladder and bananas at the top. When one monkey tries to climb, they are all sprayed with icy water. Soon, no monkey dares to climb.

Then, they replace one monkey.

The newcomer sees the bananas and goes for the ladder. The other four, fearing the punishment, immediately pull him down. He quickly learns not to climb, without ever feeling the icy water himself.

One by one, all the original monkeys are replaced.

The new group continues to attack any monkey that touches the ladder. None of them knows why. They were never sprayed. It’s just "the way things are done here."

My PhD in psychology taught me that human systems often work the same way. In our companies, we create processes for good reasons. But over time, the context is lost.

New team members are taught the process, not the purpose. They follow rules without question and enforce them on others, stifling new ideas and preventing data-driven change.

This isn't a people problem; it's a culture problem.

It's a failure of leadership to constantly re-examine the 'why' behind the 'what'.

Is your team following an invisible script they don't understand?

I see this pattern repeatedly: organisations with data scattered across disparate systems, each department working in isolation.

Sales uses their CRM. 

Finance has its ERP. 

Marketing runs campaigns through separate platforms. 

Everyone thinks they have the data they need.

But teams end up spending more time manually reconciling data across various sources than extracting value from it. You can't understand content performance or campaign effectiveness when your data lives in silos.

I worked with a client who had major system integration failures because each technology was designed only to support specific business units. No one planned how to bring together the data in a unified platform.

They were flying blind on critical business decisions. 
While competitors leveraged integrated data effectively, this organisation couldn't get a holistic view of its customer journey across touchpoints.

Their exact words: "We can't scale like this."
They were right. 

Fragmented data doesn't just create operational headaches - it directly impacts your ability to compete.

Before investing in new technology, ask yourself, “How will this integrate with what we already have?”

The Hidden Cost Data Chaos

Would you build a house on a shaky foundation?

The answer is obviously no.

However, when it comes to data quality - which is the foundation of effective analytics - I've seen clients consistently push it to the backseat. 

Why? Because their stakeholders typically see only the final results of an analysis, not the foundational work that ensures accuracy.

So, where does the data foundation start getting shaky?

It begins when your analysts are tasked with seemingly comprehensive datasets containing millions of rows of customer information, only to discover irregularities: null values, inconsistent date formats, or erroneous entries.

To cope, analysts often resort to workarounds like writing automated scripts to clean and standardise data. But automation has its limits - anomalies like "29-APR-2023" can slip through, especially in datasets with millions of rows that are impossible to manually inspect.

Faced with these challenges, analysts reluctantly make one of two choices:

1. Exclude Poor-Quality Data: Deciding that a small percentage of problematic data (5-20%) is insignificant and removing it from the dataset.

2. Include It Without Fixing Issues: Integrating the flawed data, relying on algorithms to process it as-is.

Neither choice is safe. Both undermine analysis reliability and result in:

1. Skewed Insights: Excluded data might represent critical segments like high-value customers. Missing these insights leads to poor decisions.

2. Distorted Findings: Including flawed data introduces noise that skews results, leading to inaccurate conclusions.

3. Erosion of Trust: When stakeholders discover overlooked data quality issues, confidence in both the analysis and team suffers.

In most cases, what we fail to realize is that data quality issues aren't just technical hurdles - they're strategic roadblocks that impact your entire organisation.

Don't let poor data quality hold your business back!

Many technical experts believe management is the only path to influence. There is another way to lead.

"How do I grow and have an impact in my business if I don't want to manage people?"- This is a question I get constantly from data analysts and scientists.

It’s an important question. 

The answer requires understanding the difference between managing and leading.

→ Managing often involves direct reports, performance reviews, and administrative duties. 
→ Leading is about showing direction and guiding people toward a shared goal.

You can have tremendous influence and impact in your organisation without a single direct report.

How? Lead projects. 

When you lead a high-impact project, you work with colleagues from across the business. 

You guide their time and effort. 
You own the outcome and become the authority on that solution.

This is how you build influence.

My advice for technical experts who want to grow is to focus on this path. 

Your authority will come from the value you deliver, not the title you hold.

Are your most critical tasks always deprioritised by other teams?

It’s not because they don't value your work.

It's because you haven't shown them how their small task connects to the company's biggest goal.

I faced this exact problem while working at ITV on the launch of our streaming service on a new platform. For the project manager, collecting data was a low priority against a complex software launch.

He needed to launch on time. 
My team needed tracking implemented before launch.

Instead of arguing, I used the KPI Pyramid to articulate the business-wide impact.

I explained that the reports we delivered went all the way up to the CEO. I broke down how his one small decision would ripple upwards:

1.  By not tracking this one platform...
2. ...our 'total viewer number' KPI would be incomplete...
3. ...which meant the 'total reach number' for the entire organization would be inaccurate.

Suddenly, a small technical task was framed for what it was: a CEO-level accuracy issue.

The conversation changed instantly. Tracking was prioritized and implemented before launch.

This shift wasn't about the data itself. It was about our ability to clearly articulate its impact from the ground floor to the C-suite.

Stop explaining your work. 
Start showing its impact. 
Speak in business language.

Before you agree to track another metric, ask one question: Where does it fit on the KPI pyramid?

Taplio

Dr Shorful Islam's Linkedin Analytics

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Dr Shorful Islam's Linkedin Analytics

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Dr Shorful Islam's Linkedin Analytics

's Best Posts (last 30 days)