# How to Use Data to Stay Ahead of the Curve

Data is becoming increasingly important in today's business world, and it's not going to change anytime soon.

In this blog post, we will:
- Make sure you are on board with the importance of data
- Explain how it fits in the major industries
- Guide you how to get the best data for you and to get the most out of it

We will leave you in the end with a quick overview of specialities in Data Science that might be related to your domain, and guide you in the rest of your journey.

## Does Data Really Matter?

In these times, data is more important than ever for businesses to stay ahead of the curve. And it will only get more important as time goes by.

With the right data, companies can make informed decisions that help them adapt to new trends in their market, compete with their peers and grow, gain deep knowledge of their customers, and be more efficient with their time and resources. Those who do understand and implement it in their daily operations and decisions, will rise head and shoulders above those who do not, or do it poorly.

It's no mistake that data has been called [the oil of the digital economy](https://www.wired.com/insights/2014/07/data-new-oil-digital-economy/). While the 20th century industrialisation was fuelled by the discovery, extraction, refinement and turning oil into energy, the 21st technology revolution is fuelled by data extraction, analysis, visualisations and turning data into actionable insights.
Data is today a commodity in itself: it can be sold in markets, it's governed by laws and legislations, it feeds corporations' engines of marketing, customer relationships, logistics, commerce and finances. It can even win or lose a war.

So whether you are a teenager working part-time in a mom-and-pop shop selling home baked cookies, or a quantitative developer in a top 500 fortune finance giant dealing with cutting-edge real-time trading algorithms, data does matter.

## Data Science in Business

> How data sceince applyies to businesses related to the SaaS product
> I estimate this chapter to be 250-350 words

- Ecommerce
- Health
- HR
- Logistics, Travel Industry
- Manufacturing
- Gaming
- ...

## How to Use Data

1. **Know what data you need**: Before you start collecting data, take some time to think about what kind of data would be most useful for your purposes. What questions do you want to answer? What decisions do you need to make? Knowing what you need will help you get the most out of your data.
2. **Collect data from multiple sources**: Rarely ever will all the data you need reside in a single source. Identify multiple sources, both internal and external. Your most valuable data will often be the one you collected over time. Then you can enrich it with external data as need be.
3. **Clean and organize your data**: This step is important! Make sure your data is clean and organized before you start analysing it. This will make it much easier to draw conclusions and make decisions based on your data. Remember, even the best chefs can not cook a great dish from bad ingredients, no matter the qunatity they use of it.
4. **Get a bird's eye-view of your data**: Once you have your data, take the time to carefully inspect what is at your disposal. Use [descriptive statistics](https://en.wikipedia.org/wiki/Descriptive_statistics) to summarize the data in averages, sums, counts, etc. Compute some KPIs and check their historical evolution in search of trends and patterns that strike your eye. This might also be the point where you draw visualisations of your data to be later aggregated into full-blown dashboards.
5. **Analyse data to make decisions**: All the data in the world would be worthless if it is not used to guide an action. This is the fun part, where you get to reap the fruits of your work. You need to concisely and clearly formulate a few business questions/challenges/problematics into questions that your data can answer. The answers to these will add to your arsenal of information that will allow picking the best suppliers, better recommend products to your customers, cut down in costs in your logistics chain, etc.

## Tips for Using Data

- **Formulate your business/domain questions**: Having these well-defined serves as a compass helping you navigate your road to actionable insights. An interesting dataset can be detected from the first look, but until you know exactly what you want from it, and why you want it, you won't be able to plan how to get it, and how to turn it into business decisions, an hypothesis or a finding.
- **Quality over quantity**: There's no need to collect more information than you can reasonably process and use. In fact, doing so can actually be detrimental, as it can lead you to drown in the [data lake](https://en.wikipedia.org/wiki/Data_lake) you built, or end up with a severe case of [analysis paralysis](https://en.wikipedia.org/wiki/Analysis_paralysis).
- **Build a prototype first**: Start small, with partial data, a sample, or even randomized data. Build visualisations and compute KPIs based on them. This will allow you to identify technical challenges down the road, communicate with others around the process and get their feedback early on it. Even more importantly, you can now sketch requirements and set expectations around your data.
- **Iterate on your data process**: Data and metadata will keep on changing. New sources will appear, and old ones will be deprecated. Business needs and the market will evolve with time. So a stagnant approach will leave you stuck in the past.
- **Data collaboration is key**: Make sure that you share your thoughts and plans across your team and with your clients. Get their feedback and implicate them in the process and the design of it.
- **Investigate your data, do not torture it**: “If you torture data long enough, it will confess to anything” Economist and Nobel Prize winner [Ronald Harry Coase](https://en.wikipedia.org/wiki/Ronald_Coase)
- **Do not worship data**: Your data is not an oracle, which through divine words are spoken. Your data is not a [Magic Mirror](https://en.wikipedia.org/wiki/Magic_Mirror_(Snow_White)) that will dictate your every action. Your data is a tool at your service, and should always be this way.

## Specialities in Data Science

Specialities in Data Science depend heavily on the domain of application.
Truly understanding the data and the business needs that revolve around it is what makes or breaks Data Science in the field.

Although the same mathematics and software might be involved in both Financial Quantitative Analysis and Weather Forecasting, they are two completely separate domains. So, never underestimate the domain knowledge when delving into data, or hiring a Data Scientist.

> The below will be highly adapted to the offers of the SaaS and can be used as a closing selling point

With that being said, here are some technical "pointers" that can help you better understand your data needs:

- Business intelligence and strategy: better management of resources (time, human and material), better insight into your sales, manufacturing, logistics
- AI/Machine learning: Visual Recognition, Product recommendation, NLP and sentiment analysis,
- Real Time analytics: The success of your Data Science process relies heavily on the data being dynamic, quickly updated/synchronised and swiftly notified of changes and alerts. This will come with specific technical
- ....