What is your data not doing? Get the answer you want Five lessons I've learned from Narrator co-founder Cedric Dussud -

2013. The week is Cedric Dussud's first week as the new engineering director at WeWork.

WeWork hasn't yet turned into the organization that's revolutionized how businesses run their operations. The company is managing just a few office space located in New York at the time.

It's also growing quickly.

In the first week of this month, Cedric was summoned to Adam Neumann's CEO's office.

"So this is Adam head of the department. chief, I'm also," Cedric told me. "And Adam basically says, "Hey I'm in desperate need of you to develop an automated billing system. And I need it in the next 3 weeks.'"

That's exactly the way Cedric and his staff did. Cedric's principal responsibility in the following five years was to expand the scope of the WeWork billing systems, to expand WeWork's business in its eight offices and up to 400 locations..

Through this process, Cedric learned a thing or two about handling data.

In reality, Cedric and Ahmed Elsamadisi who was an executive engineer for the data department at WeWork during the initial days, learned so much about the methods used to improve the performance of data, that they devised an completely new methods to make it simpler.

Together, they created Narrator Narrator, which is a platform for data that allows companies to conduct business intelligence operations without the expense of a massive data team.

Cedric was able to talk to me recently to talk about the main insights he learned from information in WeWork the lessons he discovered and how they contributed to the foundation of Narrator's current.

There's a myriad of ideas that stood out for me. A few of them were:

• The reason that metrics do not always add up.
• What's the next thing to do after there's no only one source that can be trusted for the truth.
• Why you should be skeptical in regards to the security of the private information that you've saved.
• If you require a professional in the field of information science can assist them (and even if you do not).

Read the full interview, or simply go through the important insights I took from our discussion below.

5 Leçons Ledric learned from Cedric's Experiment

I gained a lot from watching Cedric's interview. The following are five of my main points from the interview

1. Data Formulas are Subtly Individualized

Data science can be a huge distinct advantage in comparison to software engineering.

Within the realm that of engineering software, if have a proven technology, you can learn from what coders who have done before you.

But, since data is updated every whenever you design something entirely new, it's feasible to reproduce the code. This is why definitions may be slightly different across data reports.

Cedric employed the instance of a CEO that demanded quarterly sales reports , which were broken down into different areas.

"You usually have to begin over and create a new SQL to deal with it," Cedric explained.

"You have a feeling that, in the event you had to develop the quarterly sales report from the past, that you'll have the capability of finding a method to make use of this report to break it down by region, and maybe replicate their logic. If you are unable to comprehend the logic behind their report, you'll have to develop your own. The reasoning you use will be different in subtle ways."

If you're using this information for business purposes small changes can be detected in a short time.

"You could end up having many subtle variations in the business," Cedric told us. "And it's commonplace that the CEO will go inside the business and ask, "Why the numbers aren't being matched? What's wrong?"

2. There is no one reliable source for the truth.

These platforms with built-in analytics can provide you with a wealth of information but they don't give you any information about the flow of data between different systems in your business.

Cedric took the lead by citing advertising online.

A platform will give you specific information about the quantity of ads that were displayed, as well as their costs and other data. It can also attempt to figure out how many prospective customers have seen an advertisement prior to buying. If you're advertising across multiple platforms, there's not a single platform that can provide complete information in order to provide a complete solution, since they're not linked to each other.

"If you could sum all the orders Facebook believed it was responsible for, as well as each order Google considers it responsible for, etc this would be more than 100 percent of your orders" Cedric explained. "That's an example of how it's impossible to believe the figures. Truth isn't in any of the figures. The truth is that it is in all other cases."

3. Be sure to challenge your numbers

Cedric recommends that you not trust a company's data unless you understand exactly from where these data come from, and the way they're determined.

Also it is vital to dig deeply enough so that you fully understand and trust the facts you're dealing with to be able to make the right the company decision.

"It is just as harmful overdoing data than it can be to do this too much," he explained. "If you're thinking there's an issue and you're doing things right then you'll have to engage your mind and be thinking about the things you're doing. This is not it's not so if you've found an advanced program that will answer every question to you, and you'll say, That's cool, I'm doing an intelligent choice with my information."

4. Hire an Analyst to analyze the data you have when it is Spread across multiple Systems

There's absolutely no need for the formation of a massive analyst team from the beginning. If you're dipping far too deep into your data can lead to a naive distracting event.

"When you're in a small business that's got your head busy by the direction things are going and you're able comprehend the data you're using" stated the expert. If that's the case, there's no need to hire an analyst at the moment or at the very minimum.

But, as your technology stack expands, the numbers begin to look suspect for the CEOs and the founders.

"When you're in a position where data starts to get scattered across several sites, and you lose your ability to understand the situation."

Now is the perfect moment to have an analyst create an comprehensive plan for data -- an expert that can help assure you that you can trust your numbers once more.

5. Focus your attention on initiatives that will attract new customers.

When Cedric developed the method of billing which powers an international business such as WeWork however, the founder of WeWork would neverrecommend that SaaS founders should do similar things.

It took months of continuous work to Cedric to create a first model of WeWork's billing software used at WeWork into production, an effort that according to himisn't worth its cost in relation to smaller SaaS and software businesses.

"Work on your marketing. Find out what the requirements of your clients and explore new features, aspects that will increase your sales for your business" said the man. "Don't make up your own way for billing."

This article was originally posted here

This post was first seen on here