Skip to content

Datastrategy

We list your data, work out what it lets you decide and rank the work by how useful it is. You leave with a written plan, before you buy a tool or start a project.

Talk to an expert

For
SMEs, large companies, public bodies
Duration
From a few weeks to a few months
Team
The data specialists of Kashika Studio, in Montréal

01 When

When a data strategy is called for

All of these situations come before a decision: a tool to choose, a project to launch, a figure to defend. Before settling it, you need to know what your data really holds.
A team gathered in a bright office, around a long table and a board covered in charts
  1. Three figures for the same question

    The number of customers, a project’s margin or the hours billed change with the report you open. Everyone defends their own, and the decision waits.

  2. Data everywhere, no list

    An Excel workbook, an Access database, a management system, emails: each holds part of the information. Nobody has the full list.

  3. A project to launch on uncertain data

    A dashboard, a new ERP or an AI project is on the table. Before investing, it is worth knowing whether the data will hold up.

  4. One person knows where everything is

    They know the files, the exceptions and the fixes done by hand. A departure, a retirement or an acquisition makes writing that knowledge down urgent.

  5. Personal information that is hard to locate

    You hold data about your customers and employees. You do not always know where it sits, who can see it or how long it is kept.

  6. A tool in place, rarely used

    A reporting tool or a data warehouse was installed, but few people use it. Nobody remembers which decision it was meant to serve.

A data strategy does not necessarily lead to a large project. Sometimes it shows that your data is in better shape than you thought.

The plan then says so plainly: two things to fix, the rest to keep as it is. Nothing to buy for now.

02 Scope

What the strategy puts in writing

Each point is written down, then validated with the managers concerned. The data projects that follow build on that text.

The documents handed over

Short write-ups, reviewed with leadership and with each domain owner.

Data register
Your sources, their content, their owner and their condition, personal information flagged
Map of sources of truth
One reference per domain, its copies and what becomes of them
Priorities
The uses ranked, with the criteria behind the ranking
Rules
Owners, definitions, access by role, retention
Roadmap
The work by phase, each with a review date
Tool recommendation
Keep, adapt, replace or buy nothing, with the reasons

03 Method

From discovery to the roadmap

A data strategy has steps of its own: nothing gets programmed. Each step produces a document, validated with you before the next step. The work is done with Kashika Studio, our specialized data division.

01 Discovery

We meet the people who decide and the people who enter the data.

Leadership, finance, operations, IT: each says which decisions call for figures and which questions go unanswered. The people who enter the data know where it breaks.
Fig. 01 The questions noted during the meetings

02 Inventory

We list everything you hold.

Applications, databases, workbooks, shared folders, paper that still matters: for each source, we note what it holds, who updates it, how often and in what condition. Personal information is flagged along the way.
Fig. 02 The data register, validated with leadership

03 Priorities

We rank what your data lets you do.

Each possible use is placed by its effect on your decisions and the effort it takes: data to clean, sources to bring together, a tool to add. Useful, simple uses go first.
Fig. 03 The uses ranked by effect and effort

04 Sources of truth

We name one source per piece of data and one person who answers for it.

For each domain, one system is the reference. The others become copies fed by it, or are retired. A simple rule says where a record is created and where it is corrected.
Fig. 04 The map of the sources of truth

05 Roadmap

We put the work in order, with dates.

The priorities become phases of a few weeks, each with a visible result and a review date. The plan is yours: you carry it out with us, with your own team or with another provider.
Fig. 05 The roadmap over nine months

A project that needs software?

When systems need to be built or connected, the plan becomes the starting point of our development method. A good part of the discovery is then already done.

See our development method

04 Governance

Who answers for what, without the red tape

Data governance comes down to a few written rules, each with a name next to it. We keep the ones a company your size actually needs.
  1. One owner per domain

    Without a rule
    Everyone fixes customer records their own way. When two records disagree, nobody settles it.
    The rule
    One person answers for customers, another for projects, another for invoices. They approve changes and settle conflicts.
    For you
    A name to call when a figure looks wrong.
  2. One definition per term

    Without a rule
    “Active customer” means a purchase this year to sales, an open invoice to accounting.
    The rule
    A short, maintained glossary defines each indicator and each status. Reports refer to it.
    For you
    The same word gives the same figure, whoever calculates it.
  3. Access by role, with a log

    Without a rule
    Everyone sees everything, or each person asks for access case by case.
    The rule
    Each role sees what it needs. Key systems keep a log of who changed what.
    For you
    A clear answer when someone asks who changed a record.
  4. A written retention period

    Without a rule
    Data piles up out of caution, personal information included.
    The rule
    For each domain, we set how long data is kept, then how it is archived or destroyed.
    For you
    Lighter systems, less exposed if an incident happens.

Personal information

The register gathers what you need to know about the personal information you hold. It prepares the conversation with your legal advisor about Québec’s Law 25, without replacing it.

Where
The systems and files that contain it
Who
The roles that can access it and the reason for each access
How long
The retention period and what happens to the data when it ends
What leaves
The data passed to suppliers or external tools

05 Data and AI

Before AI, know what it will read

An assistant or an automatic extraction does no better than the data it is given. The strategy checks whether yours is ready for the task in view.

What ordered data makes possible

  • Asking about your figures in plain language and getting an answer drawn from a source of record, as in our example of conversational reporting.
  • Sorting requests or documents into your own categories.
  • Extracting the fields of an invoice or a form and filing them in the right place.

What we check first

  • The data the task needs and its real quality.
  • The person who answers for it and can validate a result.
  • The data an outside service may read, and the data that must stay with you.

What a strategy does not replace

  • An AI model is sometimes wrong, even on clean data. A human check stays in place where a mistake is costly.
  • A task is chosen for the time it frees or the errors it prevents. The model comes after.
  • A trial on your real examples says more than a plan. It can start during the strategy, on a data set that is already ready.

A task to hand to AI?

Once the task is chosen and the data checked, AI integration takes over, with a measured trial on your own examples.

See AI integration

06 After

After the hand-over, a partner who keeps the plan alive

The plan is a starting point. Your sources change, so do your questions, and data tools move fast. As your technology partner, we carry out the projects you hand us, measure what they change and advise you before each new investment. For systems already in service, monitoring and updates are covered by our support and maintenance offer.

01

Carrying out the first projects

Bringing customers into one source, cleaning a history, building the first dashboard: our data and development teams handle it if you wish, starting from the plan already validated.

02

Spotting the next weak point

Fixing one source exposes another: a duplicate that comes back, data entry that drifts, a file nobody maintains any more. At the planned reviews, we go over the register with you and adjust the order of the work.

03

Measuring the results

Are the chosen indicators moving? Do your teams rely on the figures, or go back to their own files? We check it with you before starting the next phase.

04

What AI makes possible for you

New automation and AI tools keep coming out. We tell you which ones your data already lets you use, and what would need fixing to make the most of the others.

07 Case studies

Data of every kind, put in order

Each of these projects started from the client’s files, as they arrived. The Observatoire des tout-petits receives 97 sources from ten providers, mostly in Excel. HEC Montréal’s MID platform receives 21 files every week, loaded into a database of more than 1.9 billion dated measures.

View all projects

  • Québec’s youngest children. In over 200 indicators.

    Custom software

  • HEC Montréal

    Academic research

    Thousands of artists. Measured every week.

    Custom software

  • Net-zero Canada by 2050. 151 ways to get there.

    Web application

Witify has understood our needs from day 1, educating us, guiding us and accompanying us through the evolution of our system for over 7 years now. They are the cornerstone of our technological development.

Yves Hennekens President, YHC Environnement

08 Questions

Questions and answers

What leadership teams ask us before handing over the inventory of their data.

Another question? Write to us at info@witify.io or call 1 800 334 9031.

Why does my company need a data strategy?

A data strategy helps guide and structure your company’s efforts to collect, analyze, store and use data. It ensures that your data is used optimally to support your business objectives, improve decision-making, drive innovation and maintain a competitive edge.

Without a clear data strategy, your data management efforts could be inconsistent and inefficient. They could also fail to comply with privacy and data protection laws.

What are the key elements of an effective data strategy?

An effective data strategy includes several key elements: clear objectives aligned with your business goals, an understanding of the availability and quality of existing data, and guidelines on data collection and management. It also plans the technological infrastructure needed to store and analyze data, measures to keep it secure and confidential, and a plan to develop your team’s skills.

With us, these elements take the form of short documents: the data register, the map of sources of truth, the priorities, the rules and the roadmap.

How does data strategy fit into our overall business strategy?

Data strategy must be closely integrated with your overall business strategy. Its objectives should directly support your business goals. For example, if your business objective is to improve customer service, your data strategy could include collecting and analyzing data on customer behavior and preferences to improve the customer experience.

How can we ensure that our data strategy complies with data protection laws?

It is crucial to integrate legal and ethical considerations from the outset of creating your data strategy. This can include seeking legal advice to understand applicable laws and regulations, such as Law 25 in Québec. It also takes policies and procedures to ensure data consent, confidentiality and security, along with monitoring and auditing mechanisms to ensure ongoing compliance with these policies.

Our data register shows where personal information sits, who can access it and how long it is kept. It prepares the conversation with your legal advisor without replacing it.

How long does it take to develop a data strategy?

The time needed to develop a data strategy can vary considerably, depending on the size of your company, the complexity of your operations and your specific objectives. It can range from a few weeks to several months.

It is important to note that developing a data strategy is an evolving process. It needs to be reviewed and updated regularly to remain relevant and effective.

Where do we start when information is scattered across Excel, an ERP and other tools?

With the domain that holds up the most decisions, often customers or hours. We first list the sources that touch that domain, in a single register. The others join in later phases, without stopping everything.

Excel is not the problem in itself. It becomes one when a workbook serves as a shared database: our article on Excel shows where that line is.

What is the difference between a data strategy, data governance and a BI project?

The strategy sets what to do with your data and in what order. Governance says who answers for each piece of data and under which rules: it is part of the strategy. A BI (business intelligence) project installs a tool to present the figures. It comes after.

Sometimes the plan ends with a cleaned-up workbook and a single dashboard. When more is needed, data management structures the sources and data visualization designs the dashboards.

Do we need a data strategy before starting an AI project?

Not a thick document. What you need is a short check: the data the task uses, its quality, the person who answers for it and what an outside service is allowed to read. Incomplete data gives incomplete answers, whatever the model.

Choose the task first for the time it frees or the errors it prevents. A trial on a data set that is already ready can move ahead in parallel. What follows is the work of our AI integration service.

How much does a data strategy cost?

It depends mostly on the number of sources to list, the teams to meet and the level of detail expected. Three tools and one precise question do not call for the same engagement as dozens of sources spread across several departments.

After a first meeting, we propose an engagement with its scope and its price. If development follows, our article on the cost of custom software details what drives the price of a project.

Who needs to take part on our side?

A person who sponsors the project in leadership, one contact per department involved and someone who knows your systems from the inside, in-house or at a supplier. For most participants, involvement comes down to a few targeted meetings.

If you wish, we sign a confidentiality agreement before opening your files.

Does your data deserve a plan? Let’s talk.

In 30 minutes, an expert listens to how information moves through your company and points you to the first project worth a look.

  • A 30-minute call with an expert, not a salesperson
  • A realistic budget estimate, backed by numbers
  • A concrete action plan, whether you work with us or not
Or write to us

Free. No commitment. Reply within 24 hours.

Names and figures shown in the interfaces are examples.