October 2016 - Commanders Act

Les enjeux de la sécurité de la donnée dans une entreprise data-driven

Issues Related to Information Protection within a Data-Driven Company

Written by [email protected] on October 13, 2016. Posted in Privacy.

Digital transformation started about ten years ago, if not before, depending on the definition we wish to consider. However, the subject had never been as relevant as it is today and many large companies still don’t know how they should reorganize to meet this important challenge, resulting from society’s evolution.

There are many issues associated to data in this context. Among them is the risk of data leaking, which is a major legal and financial issue.

Here are some interesting figures related to data security:

900 million data were compromised due to security breaches
88% to 90% of incidents are unintentional
In the United States, identity theft occurs once every four seconds (10 million victims)
On in 400 emails contains confidential information
One in 50 files is shared with the wrong people
One out of 10 laptops is stolen or lost
One in two USB keys contains confidential information.

Whereas hidden costs resulting from data leakage are frequently discussed, a potential, highly damaging and long lasting collateral effect is often ignored: the loss of client and user trust.

Failing to prevent a security incident inevitably harms consumers’ brand perception. What is at stake is loyalty to the latter, as it is intrinsically related to how the brand is perceived. Consequences on corporate results seem to be left aside by some companies, but the issue is being addressed increasingly and calls for enterprises to take preemptive action to guarantee data protection. According to Forrester, two to three highly ranked executives will be compelled to quit this year due to data theft.

Although all firms can be affected by data leakage to different extents (LinkedIn was pirated in 2012 and VK –Russia’s Facebook equivalent- this year), it is possible to set up a series of measures to reduce risks. One of them is systematically withdrawing obsolete tags with a TMS.

What Data is affected?

Data has become a very popular topic in conferences in recent years, and as time goes by, the differentiation between sensitive and personal information becomes blurrier. Personal data consists of information about an individual; it is associated to them through a customer code, email address or other such elements.
As for sensitive data, it consists of information related to ethnicity, political, philosophical or religious views, health or sexual preferences of an individual, amongst others. In France, Law prohibits collecting sensitive data, except when it is essential to a website’s activity, such as dating websites. Data has become a strategic matter for a large number of firms in France and worldwide, and one of the main concerns linked to it is protection.

What is a Data-Driven Company?

A data-driven company is literally a firm that is fully data-oriented. It is about a strong “data culture”, where data is not only accessible, but is at the core of strategic thinking and drives corporate action. In this context, data is a major asset for decision-making and calls for the company’s full, everyday attention.

A minimum requirement would be setting up daily dashboards adapted to every team’s needs, with a strong business focus and open to all other interested parties and containing key information they might need.

There are some departments that rely more heavily and frequently on data for decision-making than others: e-commerce sites, for example, will base their promotional strategies during sale seasons on data. But data should not be seen only as a tool to make better choices, it should be used to add value, offer better services and improve customer satisfaction.

To make the most of data, companies must stop collecting and storing it in silos, this might compel them to undergo structural and organizational changes. It entails strong collaboration between all teams and modifying the way they work.

In the current context, where data is at the core of every major strategy, observing regulations is paramount.

CNIL, Privacy and Data Protection

France’s data protection watchdog Commission Nationale de l’Informatique et des Libertés (CNIL) [National Commission on Informatics and Liberty] is in charge of protecting citizens personal information and informing them about their rights. It also issues advice to firms wishing to be compliant with new regulations; it warns and punishes non-compliant companies and organizations and anticipates future usage of personal data and information.

CNIL said that 2,800 complaints related to privacy were filed in 2015. Since its creation, the Commission has been consulted and has participated to more than 2,500 decisions and deliberations. Law 78-17 from January 6, 1978 has been modified and now comprises over 70 articles.
At the beginning of 2016, the European Union adopted a ruling on personal data with an aim at better protecting European citizens. It contains several measures and sanctions that are to enter into force in every country in the Union starting 2018.
In case a company would violate rights related to collected data, it would be subject to a fine amounting to 4% of its annual global revenue.

What Principles Must Be Observed?

1. Purpose

An organization needs to have a lawful purpose to collect personal and private information. The use it intends to make of that information must be clear and legitimate.

2. Proportion

Only necessary and relevant information to a well-defined purpose can be collected.

3. Relevance

Collected data must be relevant to the activity of the collector: a website selling socks does not require information such as gender, age, marital status and sexual preferences, whereas a dating website does.

4. Conservation Period

Data should not be stored more time than needed to serve its immediate purpose. After that, it can be stored on a different device/database.

5. Security and Confidentiality

In the United States, data theft occurs twice a day. Data protection and confidentiality are the most sensitive issues for companies, as they are compelled to guarantee secrecy and prevent intrusion, data deterioration and leakage. Security measures must obey to the nature of data and potential risks.

6. Transparency

Collecting parties must always warn users they intend to collect data and share it with third parties.
Users can decide what they share and don’t.

7. Right to Information

Users must be informed at all times about the intended use of the information they share. They have the right to modify it, control it and to approve or deny data collection and sharing.

Minimizing Risks Related to Data Protection

Data Protection Officers (DPOs) must set up the necessary protection measures to prevent data from being “damaged”, misused or accessed by anyone outside the company. Only staff or expressly authorized third parties (governmental agencies, police, etc.) having the required clearance to access and use that information should be able to do so. DPOs should also determine a reasonable amount of time to store private information, should they fail to do so, they are subject to 5 years’ imprisonment and could be fined EUR 300,000.

Minimizing risks related to data protection starts by identifying potential sources of data loss (DLP Data Loss Prevention), security breaches and assessing their importance. This entails mapping all data to protect.

In addition, data whose combination might be potentially sensitive should be coded and stored separately. Encoding keys should be modified on a regular basis and stored in external servers with secured connections.

Finally, data protection strategies must be updated quite often, as information is constantly threatened. Every time an incident occurs, an investigation must be opened to identify the source of the problem and reinforce the established security measures.

There is always a risk and no system could be 100% safe: the human factor is an indirect threat and is hardly controllable (employees are very often responsible for attacks and intrusions without knowing it). However, staff can be provided with essential guidelines and precautions to adopt in their everyday work to prevent data loss.

Who should be in Charge of Security within the Company?

The European Parliament adopted a Ruling on the Protection of Personal Data on April 27, 2016. It compels companies whose activities consist in treating data and require tracking individuals on a regular basis, to designate a Data Protection Officer (DPO) (In-house or not). The DPO must inform the company they work for about rules and obligations regarding personal data, they must speak about and train staff to comply with regulations, provide advice in terms of impact analysis and cooperate and be in constant contact with CNIL.

French firms have two years to be compliant with new Laws.

What is a DMP?

Written by [email protected] on October 13, 2016. Posted in CDP.

The ongoing digital transformation calls for major organizational changes within companies. Silo-management, which had been the norm until recently, is being challenged in favor of a user-centered global strategy. And in a user-centered organization, profile unification is indispensable, meaning that using a DMP is necessary as well.

The term DMP is being used extensively as time goes by, and not without reason: at a time when Big Data is a recurrent topic and questions surface, DMPs offer the possibility to manage and activate collected data. More important is the fact that they allow companies build data equity.

DMP means Data Management Platform; it usually refers to a platform provided as SaaS that is used to gather, centralize, generate and activate client or lead-related data. In a few words, it is a “super data base” populated by data coming from many sources and working in real time. The idea behind this type of platform is to reunite all the data bases that were managed separately and unifying them in a single data base. And even though it is a term many marketers dread, there is not a single one of them who doesn’t dream about having their own DMP.

Uses of Data Management Platforms are not limited to data storage and centralization. The most attractive feature is that they let you activate data and use it. There are plenty of different use cases.

What data is considered?

Data considered by DMPs are of many types, including browsing information (product descriptions, search queries, abandoned baskets), exposure to ads (viewed and clicked banners), offline info (TV, catalogs), CRM data (profile, CSP), voice of customer, second and third party data.
All this information allows knowing clients better but it is not its sole purpose. Every piece of it must allow action to be taken and become a trigger: recommending products, prompting basket abandoners to finish their purchase, etc.

What can you do with a DMP?

There are many DMP types. Some of them are first or third party-oriented, some others are intended for publishers or advertisers. But whatever the brand, the need remains the same: making the most of collected data.

Setting up a DMP is done through a few steps. The first one, in terms of data management, consists in collecting normalized omnichannel cross-device data on all of a client’s websites. Step two consists in unifying profiles prior to sending more complete and qualitative data to other systems. Once set-up, a DMP will significantly improve marketing operations thanks to better targeting, personalized browsing experiences and offers, more appropriate advertising investments (drop in acquisition costs thanks to optimized targeting, better use of retargeting) and a better management of sales pressure.

Data Management Platforms can be used on all levers, from emailing to in-person store visits, as well as in personalized, real-time offers displayed on a brand’s website. There are multiple possibilities:

Focusing on a clienteling approach when customers walk into the store.
Optimizing online acquisition costs by adjusting marketing pressure in favor of leads.
Using the right sales activities to increase conversions.
Limiting the appearance of chat windows and saving them for specific profiles.
Switching from statistical testing to segmented testing.
Setting up audience extensions based on discriminatory models.
Excluding clients from certain commercial operations.
Issuing follow-up messages to clients that abandoned their basket right before a conversion took place.
Displaying a pop-in the moment a user intends to leave the site to prevent them from doing so.
Simplifying payment steps for VIP clients or issuing follow-up messages for clients who have not yet issued any payment.

DMPs also let you create better segments by targeting leads whose profiles are similar to your clients’ (audience extension or lookalikes), following a successful campaign.

DMPs are also used as part of acquisition and customer loyalty development strategies: you can target users who, according to browsing data, seemed to be particularly interested in a specific product category, even though no purchase was ever made on the site.

Segmentation, targeting, service personalization through more relevant communications

Data collection on any device and channel will give you a comprehensive and detailed 360° vision of every individual. The aims of using DMPs are not limited to achieving a better segmentation, as what is sought is to take it as far as possible in terms of marketing personalization. It is all about unifying profiles, defining a segmentation strategy and making sure each segment is receiving a relevant message.

Analyzing the purchase history of a given individual is not enough to determine their needs and wishes. However, by cross-referencing browsing data, search queries and offers they reacted to, you can begin drafting comprehensive user profiles. You will have a clearer view of their interests and be able to anticipate their underlying needs. You will also improve the relevance of your communications and the performance of your advertisements while being very responsive.

Client Experience, Marketer 2.0’s Grail

Improving client knowledge continually and on a regular basis is key to optimizing one’s marketing strategy in real-time. Thanks to data collection and treatment, client knowledge is deepened.

Better client knowledge means a better user experience, provided data is correctly used, by the means of:

Relevant targeting.
Real-time personalized offers according to on-site behavior and data.
A significant cut in inappropriately targeted advertisements.
E-mailings based on searches conducted on the site.

Whatever the industry, marketers must succeed: they have to send the right message to the right audience, ensure campaigns are effective and do their best to guarantee traffic to the site converts. This entails optimizing communications, segmentation, targeting and message personalization permanently.

What is Tag Management?

Written by [email protected] on October 13, 2016. Posted in Tag Management.

Tags are used to collect and distribute data to your web analytics tools, affiliation platforms and live chat solutions; they are also used in A/B testing, advertising, social media and retargeting, among others.

The domination of “tag” technologies

Display, retargeting, marketing automation and analytics management: nearly all digital marketing solutions rely heavily on tags, and using a good Tag Management System (TMS) allows optimizing their profitability.

The number of digital marketing solution providers has skyrocketed in the past years: in 2011 there were roughly 100 editors of such solutions, while they totaled about 2000 in 2015 (the exact number is unknown since not all providers are listed).

Why do we talk about Tag Management?

The first step towards understanding tag management is defining tags. A tag is a JavaScript code snippet placed in the source code of a webpage. The tag “fires” when the page is visited or when a user interacts with html events (submitting a form, clicking and “add to cart” or other button, etc.). The tag gives the site’s owner (publisher, e-commerce site, brand, etc.) access to certain services (obtaining information, interacting with users through real-time customization of sites).

With a TMS, marketing teams can manage tags on their own and do not need any technical skill to do so. The TMS allows placing all the tags in a website into a single tag container, which is managed as a tag itself. In addition to all the organizational benefits explained in this article, the container hastens websites’ loading speed.

Using a TMS also allows to better manage tags’ lifecycles by reducing errors resulting from their complex administration: it is common to leave unused tags in a site’s source code after a campaign has ended. By using a TMS, marketing teams can be more responsive, more flexible and prevent their pages from slowing down due to the presence of obsolete tags.

Prior to the development of these systems, tag management involved several participants outside and within a company: digital marketing solutions, marketing and IT departments.

A few years ago, setting up a TMS was a complicated endeavor; it entailed a long series of procedures and specifications and allocating significant financial and human resources to the project. It was a long and expensive process.

Nowadays, Commanders Act can setup and deliver a fully operational TMS in one week thanks to their “Flash Setup” methodology.

In addition, should you cease working with a partner, there would not be any exit costs since no IT resources would be required.

Is a TMS only about tag management?

Benefits from using a TMS are not limited to tag management, they are extended to greater knowledge of your customer base. By gathering behavioral information about your clients, you get to better know and understand them. By cross-referencing data obtained through your TMS with that present in your CRM, you obtain more detailed visitor profiles. Tag management is thus more than a simple data collection tool, as it allows it to be manipulated to enrich and improve customer experience.

Tag management speeds up the implementation of unified marketing. It is about offering a coherent and harmonized global experience to each visitor, regardless of the device or channel they use and come from. A unique customer profile is created for each and every one of them and integrates data collected on all devices and channels. Tags provide a complete, multi-channel vision of client behavior and increase profitability rates for every solution used.

Tag management is also used with mobile apps and sites; some solutions provide tags for all platforms.

But tag management still has a long road ahead and will have to face new challenges in the upcoming years.

The first major challenge will be to collect a single piece of information only once and distribute it to 10 solutions at the same time. This will allow:

Increased loading speed.
Preventing conflicts between JavaScript files.
Guaranteeing data confidentiality: collected information will no longer be visible.
Guaranteeing better control over data collected and distributed to partner solutions and limiting information leaks.

data-au-service-marketing-predictif-donnees

Using Data to Predict

Written by [email protected] on October 13, 2016. Posted in Data.

By 2020 Big Data will consist of 40 zettaoctets of information (1021 octets), which is times 33 what it was in 2010, while research on the subject has increased twentyfold since 2009.

big-data-prédictif-logiciel — Évolution du nombre de recherches pour « Big Data » dans le monde de 2009 à 2016 – Google Trends

However, data has been collected since long, at a time when no one ever spoke about Big Data. But collecting data without clear purpose or strategy, as it had been done for a long time, is clearly not profitable. Fortunately, companies that collect data without knowing how to put it to good use are a rare minority. Big data has increasingly become a part of predictive marketing.

Prediction, the art of anticipating trends.

Predictive analysis consists in defining models through algorithms that use collected data. These models intend to anticipate and predict trends in consumer behavior. This allows identifying customers who are likely to become less involved with the brand and those whose needs evolve.

Prediction grants access to marketing’s Holy Grail: sending the right message to the right audience at the right time. An old marketing rule that is hard to apply without access to the right information our without analyzing it properly.

The ability to offer tailored advice at the precise moment a need is being born has significant impact on client satisfaction, and thus on their engagement towards the brand and the company’s financial results.

But prediction is only possible if collected data is good quality and varied enough. In order to have a good predictive model, the following steps must be followed:

Defining the problem(s) that the predictive model should solve.

Defining the problem(s) that the predictive model should solve.
Identifying the necessary information to set up the model.
Collecting and treating data.
Building an effective model.
Evaluating the model’s precision and effectiveness.
Using the model to solve the identified problems and issue recommendations.
Improving the model continually.

In sum, predictive analysis allows better decision-making, risk avoidance (anti-churn for instance) and differentiation thanks to an optimized client experience.

Beware of externalities and the limitations of predictive models

Cathy O’Neil does not define herself as a Data Scientist but rather as a “Data Skeptic”. Mrs. O’Neil, author of “Weapons of Math Destruction” and holder of a Ph.D. in mathematics from Harvard University, warns in her book about the potential side effects of predictive models. According to her, there are many sources of errors, including the following:

A lack of precision from collected data.
The irrelevance of used data.
Inadequate measurement of externalities.

How can it be used?

Use cases come in large numbers and many forms, depending on the industry. The first limit is creativity, but possibilities offered by predictive analysis depend on the volume, quality and relevance of the collected information.

Second party data can play a key part in predictive analysis. A car maker may have a difficult time knowing that a consumer’s family is growing, except if a partner specializing in goods for newborns provides him with that kind of information. It is then in the car maker’s best interest to propose to its potential customer a personalized offer, adapted to their new situation, by serving personalized advertisements or modifying the website’s homepage according to the user’s profile, and highlighting a vehicle they might considering purchasing.

Predictive analysis also lets mobile operators identify users who constantly make more phone calls than those allowed by their mobile phone plan, allowing the companies to better target users with new adapted plans.

Predicting entails serving content adapted to each and every consumer’s particular situation, not only making efforts to sell a product or service among users for whom a clear need has surfaced. In this environment, well-performed predictive analysis represents a strong competitive advantage for companies that engage in such practice. It is becoming a key element in terms of decision-making.

Is predictive marketing for experts only?

According to a Forrester study, predictive models have become more precise and effective in recent years – even within firms that lack skills related to that field – thanks to new tools that are easier to use.

Several companies may believe that data analysis requires hiring a Data Scientist, but such a belief is not all that accurate and they should focus on training the human resources they currently have. According to Forrester, the larger the number of employees working with predictive analysis, the better. Including those lacking specific statistical skills.

connaissance-client-first-second-third-party-data

Understanding Customers Better Through First, Second and Third Party Data

Written by [email protected] on October 13, 2016. Posted in Data.

Nowadays, Big Data is essential to understanding customers better. It is made of different data categories, including first, second and third party. What do they consist of? What are their main differences? Are they all necessary? Which ones should be prioritized?
Here are our answers to your questions.

The main difference between the three types of data is the collection level and proprietorship; all three types are complementary and together help having clearer and more precise customer profiles.

What is First Party Data?

First party data comprises data collected directly by a site’s owner: it includes browsing and behavioral information, as well as other details collected through forms, search queries, etc. First party data also consists of information gathered through other sources such as CRMs, offline databases, contests, among others. This data is anonymized.

The main objective of first party data is understanding consumers’ intentions and interests. E-commerce sites can make relevant product suggestions to users and increase conversion rates thanks to this data category. Nevertheless, and even though first party data is essential to having detailed customer profiles, it is not enough. This is where second and third party data come into play: they complete first party data with large volumes of additional information.

What is Second Party Data?

Second party data is basically first party data belonging to a third party who shares it with you as part of a collaboration agreement. The level to which second party data will complement your own first party data, will depend on the type of partner you work with. Second party data allows obtaining a more complete user profile, mostly in terms of interests and other aspects that are not necessarily covered by first party data.

Second party data, also known as collaborative data, is useful as first party data and comes in large volumes, just like third party data.

What is Third Party Data?

Third party data is made of information that is collected and sold (or lent, to be precise) by data providers. It is widely believed that it is of lesser quality than first and second party data, but there are exceptions: Facebook, for instance, has complete and continually reliable information about its users. It combines data it collects as first party and data it obtains through advertisers.

The use of third party data provides a wide choice of possibilities, since the vision it gives about user interests is much broader than that offered by the other data categories. By combining them, user profiles become more precise.

Why Don’t We Just Use First Party Data?

First party data has many benefits: it is very affordable (free most of the time), easy to collect, seems as the only reliable information to many advertisers … Nevertheless, it is not sufficient to get to know your clients if used on its own, especially in terms of acquisition.

Not all industries can collect this type of data in the same way: banking institutions, connected object manufacturers or e-commerce sites have access to larger volumes of first party data than companies from the automotive, appliance or other activities that are not in direct contact with consumers.

Companies in such industries are very likely to make use of second party data, and in case building partnerships would not be possible, they still have third party data as an option, which they could trade services for (free media spots, services, etc.).

Without second and third party data, acquisition campaigns can hardly be personalized – except through Facebook and Google, which propose to inject your data to target defined segments. The conversion rates of personalized messages are much higher than those of generalized and impersonal communications. Acquisition costs can double if second and third party data are not exploited.

First party data also plays a crucial part in terms of acquisition, as it provides vital information related to customer knowledge. This information is treated to send relevant messages to a given audience amid an acquisition campaign through the use of statistical twins: people whose information reveals similar profiles.

Solutions

Products

Resources

Compagny

Archives

Categories

Month: October 2016