R. c. Société Telus Communications est un autre exemple jurisprudentiel intéressant de l’application du droit général à la protection contre les fouilles, les perquisitions et les saisies abusives garanti par l’article 8 de la Charte canadienne en fonction du progrès technologique.

La Cour suprême a dû se prononcer sur la validité d’un mandat général obtenu par la police pour mettre la main, de manière prospective, sur des messages textes (« textos ») auprès d’un fournisseur de services de télécommunications, soit Telus. Il s’agit d’un appel d’une décision de la Cour supérieure de justice de l’Ontario.

À titre de précision préliminaire, la communication « prospective » vise des informations futures, en l’occurrence des textos qui sont envoyés après l’émission du mandat ou de l’autorisation.

Ce ne fut visiblement pas une mince affaire à trancher en ce qu’elle a divisé la Cour sur plusieurs plans. Cinq (5) juges, soit la majorité du panel en comptant sept (7), se sont entendus pour annuler le mandat général, mais ont émis des motifs qui divergent :

• Trois (3) juges ont déterminé qu’une « autorisation pour écoute électronique » en vertu de la partie VI du Code criminel était essentielle, comme c’est le cas pour intercepter des conversations téléphoniques verbales, entre autres pour assurer une meilleure protection de la vie privée des individus qui en font l’objet. La juge Abella est l’auteur des motifs relatifs à cette opinion et les juges Lebel et Fish y ont adhéré;
• Les juges Moldaver et Karakatsanis ont préféré ne pas se prononcer sur la question de savoir si une autorisation pour écoute électronique était nécessaire dans les circonstances. Il ont par ailleurs exprimé leur accord avec les motifs de la juge Abella quant au fait que la Couronne n’avait pas comblé son fardeau d’établir qu’il n’existait pas, ailleurs dans le Code criminel ou les lois fédérales, de disposition permettant d’autoriser la perquisition desdits textos autrement que par un mandat général.

Les deux (2) autres juges, soit les juges Cromwell et McLachlin, sont dissidents. Ils étaient d’opinion que le mandat général avait été obtenu par la police en conformité avec les exigences de la loi.

Théoriquement, on ne peut dire que cet arrêt crée une règle faisant l’état du droit quant à la nécessité d’une autorisation pour écoute électronique dans des circonstances analogues. Ce ne sont que trois (3) juges sur sept (7), soit un de moins que la majorité du panel, qui sont de cet avis.

Par contre, en pratique, on pourrait s’attendre à ce que l’obtention d’une autorisation pour écoute électronique devienne désormais la norme chez les autorités policières dans de telles circonstances. En effet, on peut penser qu’il leur serait plus prudent de faire ainsi, notamment puisque la procédure d’obtention de cette autorisation ainsi que ses modalités confèrent aux usagers de la téléphonie portable de meilleures garanties de protection de leur droit à la vie privée. Ce type d’autorisation serait donc moins susceptible de faire l’objet d’une contestation.

Le présent billet focalise son analyse sur les motifs de la juge Abella, considérant qu’ils sont ceux qui, à notre avis, auront les plus grandes conséquences pratiques.

 Contexte factuel

En premier lieu, il faut préciser que, contrairement à plusieurs autres fournisseurs de service de télécommunications, Telus enregistre pour une durée d’environ 30 jours, dans son système informatique, copie des textos échangés par ses clients notamment pour améliorer la résolution de problèmes techniques et pour mieux gérer les plaintes de sa clientèle.

Ainsi, lorsque les policiers s’intéressent aux textos échangés par les clients de Telus, ils demandent l’émission d’un mandat ou d’une autorisation pour mettre la main sur la copie des textos enregistrés dans ce système informatique.

Dans les dernières années, Telus a reçu des dizaines de milliers de mandats de perquisition et d’ordonnances de communication de textos, et ce, à raison d’environ 350 par mois. Parmi ceux-ci, à l’époque où le litige a été déclaré, Telus n’en avait obtenu que six (6) sous forme de mandats généraux. Tous les autres étaient sous forme d’ordonnances de communication rétrospective ou bien, d’autorisations pour écoute électronique visant les communications futures.

Le mandat général en litige requérait à Telus de fournir à la police quotidiennement, tous les textos envoyés par deux clients de Telus, en plus de l’information identifiant les individus à qui ces textos étaient envoyés et/ou qui les avaient reçus.

Quoiqu’il était plus simple et rapide pour les policiers d’obtenir un mandat général, nous comprenons qu’il était beaucoup plus laborieux et coûteux pour Telus de se conformer à celui-ci en comparaison avec une autorisation pour écoute électronique.

Telus a fait de ce dossier une cause type en demandant l’annulation du mandat général au motif qu’une autorisation d’écoute électronique en vertu de la partie VI du Code criminel était plutôt requise pour perquisitionner les textos de manière prospective, notamment afin que ses client puissent jouir de meilleures garanties relatives à leur droit à la vie privée.

Telus a été débouté en première instance en mars 2011. La Cour supérieure de Justice de l’Ontario avait statué que le mandat général était valide pour la communication de textos de manière prospective.

Par contre, la Cour suprême a renversé le jugement de première instance et a annulé le mandat général.

Le mandat général versus l’autorisation pour écoute électronique

Selon l’interprétation de la majorité à la Cour suprême, un mandat général a un caractère résiduel et ne peut être décerné si une autre autorisation est disponible dans la loi afin que les policiers puissent obtenir l’information convoitée. Ce mandat ne peut donc être obtenu en premier recours et la Couronne a le fardeau de prouver qu’aucune autre autorisation n’existe dans la loi permettant à la police d’arriver à ses fins.

Essentiellement, il est possible d’affirmer que, pour la police, les conditions d’obtention et les modalités d’un mandat général sont moins exigeantes et que celles énoncées à la partie VI du Code criminel intitulée « Atteinte à la vie privée » pour l’obtention d’une autorisation pour écoute électronique. Ce dernier type d’autorisation permet d’intercepter des communications privées.

Selon les motifs de la juge Abella, les dispositions de la partie VI assurent une protection accrue du droit à la vie privée consacré à l’article 8 de la Charte canadienne des droits et libertés. Selon celle-ci, les policiers auraient dû obtenir une autorisation pour écoute électronique. Les motifs au soutien de son opinion peuvent être synthétisés comme suit :

• Similitudes entre les conversations téléphoniques et les communications via textos : La juge Abella nous fait remarquer que « [l]a messagerie texte est, essentiellement, une conversation téléphonique. La seule distinction entre la messagerie texte et les communications orales traditionnelles réside dans le processus de transmission. Cette distinction ne devrait pas avoir pour effet de priver les messages textes des mesures de protection des communications privées auxquelles ces messages ont droit sous le régime de la partie VI. Les différences techniques intrinsèques des nouvelles technologies ne devraient pas déterminer l’étendue de la protection accordée aux communications privées. » La juge Abella poursuit en mentionnant que ces textos sont échangés dans un contexte créant une atteinte raisonnable de vie privée et constituent donc des communications privées au sens de la partie VI du Code criminel.
• Interprétation de l’expression « intercepter une communication privée » : Tel qu’évoqué précédemment, la partie VI du Code criminel permet d’intercepter des communications privées.

La juge Abella est d’avis que dans l’hypothèse fictive où Telus ne stockait pas les textos dans son système informatique, il faudrait littéralement « intercepter » les textos simultanément à la conversation et nécessairement, une autorisation pour écoute électronique serait de mise.

Par contre, ici, la réalité diffère : la police voulait communication des textos non pas simultanément à la conversation, mais plutôt après que ceux-ci aient été stockés par Telus dans une base de données. De ce fait, pouvait-on dire qu’ils étaient « interceptés » ou qu’ils étaient plutôt « divulgués, utilisés ou conservés » après la survenance de la conversation ? La juge Abella répond qu’ils sont, à son avis, interceptés. Celle-ci préconise une interprétation large et libérale de l’expression « intercepter une communication privée » et rejette une interprétation stricte et formaliste. Selon elle, l’interception ne doit pas nécessairement avoir lieu en temps réel, simultanément à la conversation.

La juge Abella précise ce qui suit: « L’emploi du mot « intercepter » implique que la prise de connaissance de la communication privée se fait au cours du processus de transmission. À mon avis ce processus englobe toutes les activités du fournisseur de services qui sont nécessaires ou accessoires à la fourniture du service de communication. La prise de connaissance de la substance d’une communication privée se trouvant dans un ordinateur exploité par un fournisseur de services de télécommunication ferait, en conséquence, partie de ce processus. »

Cette interprétation large et libérale fait donc en sorte, selon la juge Abella, que la partie VI soit applicable dans les circonstances pour la communication des textos de manière prospective, laquelle assure une meilleure protection de la vie privée des individus qui en font l’objet.

Sur le même pied d’égalité

Dans la mesure où les autorités policières ne procèdent dorénavant qu’avec des autorisations pour écoute électronique dans des circonstances similaires, la vie privée des clients de fournisseurs qui stockent les textos dans un système informatique sera protégée de la même façon que celle des clients de fournisseurs qui ne stockent pas ces données.

Cela éviterait donc une disparité de traitement, laquelle pourrait potentiellement créer des insatisfactions chez la clientèle des fournisseurs qui choisissent de stocker les textos.

Cela ferait également en sorte que le fait pour un fournisseur de stocker les textos, notamment afin d’améliorer la résolution de problèmes techniques et pour mieux gérer les plaintes de sa clientèle, n’aurait pas pour effet de le désavantager du point de vue des coûts et ressources requis pour se conformer aux autorisations obtenues par la police pour la communication de textos de manière prospective.

Information disseminated through social media platforms such as Facebook and LinkedIn is of growing utility in litigation matters. Evidence obtained from social media accounts by way of discovery preservation and production orders has significantly strengthened the positions of litigating parties. This should come as no surprise as individuals routinely “post” messages, thoughts, pictures and experiences on these platforms, leaving a wake of evidence in the process.

There has been marked development in this area of law in Canadian jurisprudence. To date, Courts and Tribunals have, among other things, ordered the preservation and production of entire social media accounts, dismissed wrongful dismissal claims based in part on the disparaging nature of comments posted online, and considered social media evidence against claims of the loss of enjoyment of life and the inability to work. However, the Courts are cognizant of the private nature of this information and have been careful to balance the probative value of this evidence against the privacy interest of the social media user. Generally, the Courts have resolved this tension by making a determination of how “private” the social media account is through the application of a number of factual indicia. As this area of law develops, Canadian Courts and Tribunals will no doubt find increasing utility in the evidentiary value such information provides.

When Social Media is Producible

Canadian jurisprudence is clear that social media accounts are considered documents that must be produced if they contain relevant and material information. This principle manifests itself most acutely in personal injury claims where the “social” aspect of social media speaks directly to the claimant’s loss of enjoyment of life. For example, social media such as pictures of claimants engaged in recreational activities are often admitted as evidence relevant to demonstrating the claimant’s enjoyment of life or ability to work.

However, the probative value of social media is not limited to personal injury claims. In one instance, disparaging comments made against an employer by an employee on a blog were sufficient to dismiss the employee’s claim for wrongful dismissal. In another, a former employer was able to obtain an Anton Piller order to seize, among other things, the LinkedIn account of an employee that they claimed had breached confidentially restrictive covenants in an employment contract.

Many Courts have inferred the likely existence of relevant documents on a limited-access account from the nature of the social media service. Some Courts have denied such an inference, finding instead that private information on limited-access accounts is producible only when publicly available information infers the existence of relevant material held privately. Interestingly, some Courts have cautioned that such accounts are likely to contain a degree of “puffery” that must be taken into consideration.

In determining the privacy interest of the social media user, many Courts point to factual indicia of privacy surrounding the account itself. Many social media accounts are of limited access, containing internal controls, set at the discretion of the account holder, that limit the viewable content and the discoverability of the account. Courts have also considered the number of individuals able to view the social media account. In one example, the claimant’s Facebook account was viewable by 200 Facebook “friends”, only 5 of which were described as “close friends”. The Court concluded that this wide audience mitigated against privacy and the Facebook account had to be produced.

The following table contains a list, with noteworthy considerations, of select reported Canadian cases requiring the production of social media accounts:

When Social Media is not Producible

Canadian Courts have clarified instances when social media accounts are not producible. It is clear that where the social media accounts are not relevant and material to the pleadings, they need not be produced. Production is also not required in instances where the privacy interest of the account holder outweighs the probative value of the evidence. In one example, a claimant’s Facebook account was viewable by only 67 Facebook “friends” with strict privacy settings. The Court implied that this mitigated in favour of privacy, and the account was not producible. In a second example, the Court noted that permitting access to 139 “friends” operated to exclude approximately 1 billion Facebook users, showing a privacy interest that prevented production.

The following table contains a list, with noteworthy considerations, of select reported Canadian cases not requiring the production of social media accounts:

Conclusion

When hearing applications for the production and preservation of social media accounts, the Court is being asked to engage in a delicate balance between the privacy of the individual and the probative value of the evidence. Two factors that clearly erode an account holder’s privacy interest include relaxed privacy settings and the existence of relevant information available on public portions of the account that allows the inference of the probable existence of information held privately.

As shown above, a third factor, being a large number of “friends” able to view the social media page, has proven unclear. However, it is likely that Courts are more concerned with the nature of the account holder’s relationship with the audience, as opposed to its size. When making an application for the production or preservation of social media accounts, it may be more appropriate to frame the argument as being about the nature of the “friendship” as opposed to the quantum. The more “public” the audience is, the less likely there exists a legitimate privacy interest.

* Jeremy Busch-Howell is an Articling Student at McCarthy Tétrault.

Can employees be trusted to search their own records in response to an e discovery request? What documentation is required to demonstrate “search accuracy” in self-collection? What if the custodians get it “wrong”?

These are some of the questions addressed in the Opinion and Order filed July 13, 2012 in National Day Laborer Organizing Network v. The United States Immigration and Customs Enforcement Agency in the United States District Court, Southern District of New York. The matter concerns a 21-page request under the Freedom of Information Act for information from a variety of federal agencies and departments about the Secure Communities, an aspect of National Immigration Policy.

In response to the plaintiffs’ request, the defendants claimed that they have spent “thousands of hours and hundreds of thousands of dollars.” The adequacy of those searches was the subject of the latest cross-motion in this case.

Although the decision is highly fact-specific, it nevertheless contains a summary of U.S. case law on the adequacy of self-collection and employee-based searching for electronic documents in response to information requests.

In determining whether the searches were reasonable, Judge Shira Scheindlin considered whether:

• the defendants’ decision to exclude certain custodians was reasonable
• the locations, places and systems searched were reasonable
• a search of any former employees’ systems was required
• the search terms were adequate and reasonably tested
• the precise instructions given by custodians to their computers were reasonable
• the search terms as combined and deployed were reasonable
• the description of the search terms and search methodology were adequate

Search Documentation – Search Methodology and Search Term Description

The judge observed that to determine the adequacy of a custodian’s self-collection, the court must understand how the custodians developed and implemented searches, because the custodians “never actually look at the universe of documents and rely on search terms” to produce a subset of potentially responsive records that are then examined for responsiveness. Additionally, “[i]n order to determine adequacy, it is not enough to know the search terms. The method in which they are combined and deployed is essential to the inquiry” to understand whether the searches are reasonable.

According to the judge, custodians must provide a description of the searches because the Act places a burden on the searching party to establish the adequacy of searches. This can be accomplished by having the searching party record and report on the search terms used, how these were combined, and whether the searches searched the full text of documents. The judge also noted that most custodians cannot be trusted to run effective searches because designing legally sufficient electronic searches is not part of the custodian’s daily responsibilities. She distinguished searching for an answer on Google as being very different from searching for an “all responsive documents” in an e-discovery context.

The judge cited evidence that “key word searching is not nearly as effective at identifying relevant information as many lawyers would like to believe.” She cited previous decisions which speak to “a need for careful thought, quality control, testing, and cooperation with opposing counsel in designing search terms or ‘key words’ to be used to produce e mails or other electronically stored information.”

Court Evaluation

If custodians do keep track of and report on the search terms they have used, how should a court evaluate their adequacy? The judge acknowledged her inability to do so in this case but referred to the plaintiffs’ search expert’s evaluation of the inadequacy of search terms.

As the court could not provide the parties with specific guidance on the adequacy of searches, it directed the parties to meet and confer and devise a mutually agreeable scheme to supplement the material already provided by the defendants under the information request. She also suggested that the parties examine emerging technologies, including computer assisted or predictive coding technologies in order to complete the remainder of the searches under the request.

Our Experience

In our experience, most custodians cannot be left to conduct searches without the involvement of attorneys or other qualified personnel experienced in e discovery initiatives. We note moreover that the cost of re-doing a search is more frequently larger than the cost of correctly identifying potentially responsive documents at the outset of the e discovery request.

Search is challenging because the adequacy of a search can depend entirely on context and available IT technologies and the characteristics of the information being searched. Employees frequently do not have enough information to evaluate the context of requested searches.

Furthermore, the search opportunities that may be done initially within a company’s information systems may not be the same and may be inferior to searches done following a collection of potentially responsive documents and their transfer to a system specifically designed to assist in the identification and review of potentially responsive documents.

Our own experience suggests that companies that actively engage counsel and e discovery expertise at the earliest stage following a request for information will save costs and reduce the production and response burden when responding to e-discovery requests.

What are the costs associated with different phases of e-discovery production? How are these costs distributed across internal and external sources of labour, resources, and services? And how can these costs be reduced without compromising quality? The Rand Institute for Civil Justice’s April monograph, “Where the Money Goes: Understanding Litigant Expenditures for Producing Electronic Discovery” explores these issues, highlights the main challenges related to preserving electronic information and provides some recommendations on how to address complaints of excessive costs and uncertainty.

In the Rand study, the researchers collected data from eight major corporations across 57 cases and interviewed key legal personnel from these companies. The companies represented in the study include one in each of the communications, electronics, energy, household care product and insurance fields, as well as three from the pharmaceutical/biotechnology/medical device field.

Although the researchers cautioned that their case study approach did not permit them to draw generalizations that would apply to all corporate litigants or all discovery productions, their findings still provide a detailed account of the resources required by a diverse set of very large companies to comply with e-discovery requests.

Findings

The researchers found:

• The review of documents for relevance, responsiveness and privilege was the major cost component one-discovery production, typically about 73%
• Collection consumed about 8% of e-discovery expenditures
• Processing costs consumed 19%

These costs broke down into the following, by source:

• Expenditures for the services of outside counsel consumed 70% of the total e-discovery production costs
• Internal expenditures were 4% of the total (even adjusting for under-reporting)
• Vendor expenditures were 26%

In terms of potential avenues for cost reduction, the researchers observed that although significant reduction in current labour costs is unlikely and that increasing the rate of review has its limitations, they were cautiously optimistic that computer-assisted categorization of documents may assist in reducing costs:

Finally, some important generalizations emerged regarding the main challenges faced by corporate counsel related to preserving electronic information in anticipation of litigation. In particular:

• Companies are not tracking the costs of preservation
• Preservation expenditures are significant
• There are complaints about the absence of clear legal authority about what strategies are defensible for preservation

Recommendations

To address the complaints of excessive costs and uncertainty, the researchers recommended:

• Adopt computer-assisted document categorization (predictive coding) to reduce the costs of review in large scale e-discovery efforts
• Improve tracking of costs of production and preservation
• Bring certainty to legal authority concerning preservation

Our Experience

Our observations are consistent with the conclusions in the Rand Report. We have found that, historically, the greatest proportion of costs was attributable to document review, even where this was contracted out to external document review entities. However, recently we have seen some developments that have lowered the overall cost of electronic document production including the cost of review.

• Companies have invested in information systems and record retention protocols, which result in the rapid and effective identification of potentially relevant documentation. As a consequence, there are fewer examples of over-collection, and the corresponding costs of processing and review
• Technology for identifying relevant documents at source has improved dramatically in the past four years. This technology now also permits greater, and cheaper, de-duplication of files, reducing overall volume for review
• The increase use of technology within the document review phase is increasing the speed and accuracy of the review, while assisting in the identification of wholly irrelevant documents without the need for human review

We remain cautiously optimistic that computer-assisted review will lower the overall cost of pre-disclosure document review. However, computer-assisted review is not a substitution for human involvement in process design, including for the careful direction in the preservation and collection of records, the analysis of the issues involved in a dispute, and a co-ordination of document management with the overall litigation strategy. A number of factors can improve outcomes and lower costs, and each of these should be evaluated and applied in every case.

With the massive number of emails and other electronic documents generated today on a daily basis, organizations face tough challenges in preserving documents in the face of litigation. In Voom HD Holdings LLC v. EchoStar Satellite LLC, Index No. 600292/08, 2012 NY Slip Op 00658 (January 31, 2012), the New York Appellate Division provides some timely guidance on what steps parties should take to preserve documents when faced with the prospect of litigation.

The case involved a contractual dispute where the parties had been engaged in contentious discussions for six months prior to the filing of the lawsuit. The defendant did not implement a litigation hold until four days after the filing.

In the decision, the court discussed two important points.

Key Takeaway

While Voom HD Holdings is not a binding precedent in Canada, its teachings regarding preservation of electronic documents should be reviewed by organisations faced with looming litigation in Canada or abroad.

* with the assistance of Mike Yuzdepski

The production of electronic records is a central concern in all forms of litigation and government and regulatory investigations and proceedings. Strategies and trial outcomes can turn on the burden, costs and potential pitfalls surrounding the management of this process.

McCarthy Tétrault LLP has launched “Understanding Canadian e-Discovery,” a video that introduces and demystifies the issues that face clients when they meet a statutory or regulatory obligation to locate, review and produce electronic records. The video is relevant to wide range of litigation, investigation, compliance and business matters.

A recent survey on information retention and e-discovery practices, undertaken by a large IT service provider, yielded some surprising results.

The purpose of the survey was to better understand how companies are responding to litigation and regulatory requests for information despite an increasing amount of data from a growing number of sources. The survey also asked whether companies had implemented formal information retention and e-discovery practices, and the impact of doing so.

The survey uncovered the following facts:

• E-mail is no longer the most commonly requested category of record
• Instead, both loose files and information from databases outrank e-mail in production requests
• There has been a rapid rise in demands for access to both social media and mobile phone text messages
• Data in SharePoint sites is becoming more commonly requested

 Companies are feeling the impact e-discovery requests:

• On average, respondents said they had to respond to legal, compliance or regulatory requests for electronically stored information 63 times in the past year
• To find this information, IT staff needed an average of 66 hours.  That’s more than 4,000 hours per year or the equivalent of two person-years

There are wide variations between companies in information retention practices:

• Nearly half of respondents do not have any formal document retention program
• Top-tier companies that closely followed best practices were:
  • 81% more likely to have a retention plan in place
  • 63% more likely to implement the automation of legal hold
  • much less likely to follow poor information management practices, such as performing legal holds in their backup systems (vs an archive or offline storage tool)

Companies who responded to the survey reported that in cases where they were either late, partially or completely unsuccessful in responding to an information request, they experienced the following consequences as a result:

• 42% suffered damage to the enterprise reputation or embarrassment
• 41% suffered fines
• 38% suffered a compromised legal position
• 28% were sanctioned by the court
• 26% had a reduced ability to make decisions in a timely fashion
• 25% reported the delay or lack of success raised their profile as a potential litigation target

The recommendations that emerge from this survey reinforce those that are commonly cited as records management and e-discovery best practices:

1. Create and implement a records management (RM) program. Begin with a formal plan, reviewed by counsel for compliance with applicable legislation, and refine that plan over time to address specific laws and regulations governing the retention and availability of information specific to your organization or industry and new and emerging technologies.  Without a formal records retention plan, employees will not know what they must retain or what, if anything, they can delete.  Failure to delete information which may by law be deleted risks over-retention, and could create additional risk and unnecessary e-discovery expense.
2. Follow the RM program, including any provisions for deleting electronically stored information (ESI).  Most retention plans will allow companies to delete certain ephemeral and non-business information according to a record retention schedule defined by the RM program.  If yours is one of the almost 20% of organizations that retains archived data forever, your organization may be at risk for increased storage, litigation exposure and e-discovery costs.
3. Think about emerging technologies such as social media, cloud data, instant messaging and structured data systems and include these in your RM program. E-discovery does not just involve loose files and e-mails. A data retention policy identifies where all electronically stored information resides company-wide so that ESI sources do not go unrecognized.  New e-discovery tools may be required to defensibly collect and process this ESI for review.
4. Use automated legal hold processes and solutions. Automated legal hold processes can be effective in communicating the importance of a given legal hold notice, and to track responses and issue reminders to affected custodians.  A manual system for legal holds may break down across many custodians or systems of data.
5. Perform litigation readiness exercises.  Companies that practice “fire drills” and which test e-discovery and record retention systems are in a much better position to withstand challenges to their internal processes.

Here are five key steps you should take once you are sued, or know you may be sued:

1. Make someone responsible. Make someone in your organization responsible for taking the following steps. Make sure they keep a record of everything that they do.  Pick someone with enough authority and skills to actually follow through on implementing the steps below.
2. Stop routine destruction. Some computer applications and systems are set up to automatically purge or destroy documents at preselected intervals. E-mail systems and document management systems, in particular, are at risk of purging and data destruction. Make sure you inquire as to what the destruction and deletion procedures are, and stop the routine deletion, if necessary.  Also consider taking steps to stop destruction or recycling of routine backup tapes. This step can have significant consequences. You may need to buy additional storage or backup media. Consider too the configuration of your active directory, and whether there are any profiles which are about to expire; make sure these expiry dates are extended if necessary. Finally, consider routine hardware or media replacement schedules as you can end up accidentally deleting data by replacing someone’s laptop.   Before you resume routine deletion or destruction activities, consider consulting with litigation or e-discovery counsel.
3. Develop a data map.   In order to figure out which of your electronic records might be needed for the litigation, you’ll need to know what data you have and where it is.  How are your IT systems configured? Do people connect to the network, or are they remote or offline? Where is the data stored? Is it local (on machines) or networked? Is data in archive systems or backup systems? Make a list of all the places your organization stores data, what might be stored where, and in what format. In particular, make a note of places where data may be volatile or at risk, including systems that might store data that can be easily overwritten, including in databases and on mobile devices, or data that might be at risk because individual people and machines or systems need to be moved or are mobile.  Take any necessary steps to secure that data if it is important to the litigation.
4. Identify Custodians. Custodians can be people or systems that probably know something or have information about the subject matter of the litigation. Make a list of them. This will be important later on. Do this exercise early, and make sure you also consider how information flows through these people. For example, does a person send correspondence out through an assistant or direct or indirect report? If so, include those people on the custodian list too. Custodians can also be systems, including shared network drives, so make sure you include such places on a custodian list.
5. Issue a Litigation Hold notice. Preserve documents, and make sure key people in the organization know of this obligation. If you cannot identify custodians or systems, you may want to send the notice out to a broad range of people or the whole organization. If you have completed the data map and the custodian list, you can target the notice to those people. Advise people what the litigation is about and the scope of what you need them to retain. If you don’t know what documents they need to retain, get legal advice on this issue. It is also helpful to set up a system to audit people’s compliance with the notice, so that you can show, later on, that the notice was sent out, received, read and understood, and complied with. There are web-based systems that can help with this, but an effective system can also be set up with response buttons on an e-mail system and an excel spreadsheet. As a final note, make sure that custodians preserve paper.

Preservation is an important early step in preparing for litigation.  The earlier you know what information you have, the easier it will be to educate your lawyers about your position, and develop a streamlined system to collect documents that might be required in the litigation. Remember – just because you preserve it doesn’t necessarily mean you will collect and review it; however, you can’t produce what you haven’t preserved.