The Complete Guide to Contract Drafting Technology

Edition 2024

Download as PDF file.
Image of a large library

Getting to a first draft

As soon as an idea has formed in the parties’ heads about the main principles of the contract, one of them will “hold the pen” and provide a first draft.

That first draft can be created with the help of either a template or a precedent. Many lawyers use these terms interchangeably, but there is an important difference between the two that determines the readiness of an organisation to engage with different kinds of legal drafting technology.

  • Template: a document that contains standardized provisions, placeholders where variable information needs to be, and a varying degree of optionality – think optional paragraphs, fallback clauses, and conditional annexes. This information is usually embedded into the template with placeholders, comments, and drafting notes. Depending on how complex the subject matter is and how much additional guidance is offered by the template, you can expect this base document to get you within 60-80% of where you need to be.  
Example of a well-developed template
Example of a well-developed template
  • Precedent: a document that was previously drafted for another client or file, and which more or less aligns with the draft the lawyer is about to produce now. Unlike templates, precedents are not curated. You just select a document from the DMS system, email inbox, hard drive, etc. This doesn’t require any setup or maintenance, but there are also crucial limitations:
    • the pool of available starting documents is limited to whatever you can remember (or your colleagues remember, if you decide to go knocking on a few doors), and
    • you need to expend a lot more effort to extricate variable information, clauses, and legal/commercial arrangements specific to the parties for which the contract was originally drafted, deal-specific information, etc.

As a result, a precedent can get you within 20-70% of where you need to be.

This is where we arrive at the “choose-your-own-adventure” part of this guide:

  • Do you create your first drafts on the basis of precedents and have no need or desire to invest in the development of standardized templates? Then go to the section on Skeleton automation
  • Do you (want to) invest in building templates to increase the speed and quality of your first drafts? Then skip to the section on Document automation.

Note: recently, a third category of first draft has emerged: AI-generated documents.

It’s an enticing idea – just describe the document you want, and the AI will generate the entire document for you. This works reasonably well for small and simple documents, but our own observations and those of industry experts, point out that this approach is simply too flawed for anything substantial:

  • Review Time: AI-generated drafts are unpredictable and inconsistent, leading to varying results that require significant lawyer time to review and correct.
  • Knowledge Retention: Generative AI fails to capture knowledge gained in the drafting process by building and refining templates or precedents that can be reused.
  • Risk: Relying on AI introduces risks related to accuracy, reliability, and completeness. Of course, these can be mitigated, but they require significant manual review every time a document is drafted.

1. Skeleton automation

Document automation can cause tremendous boosts in drafting efficiency, but it also features a setup cost and a maintenance cost. It should therefore be used primarily on those types of documents that legal teams draft with sufficient frequency.

Even where it makes total sense to leverage document automation, many organisations – small in-house legal teams and law firms without knowledge management professionals in particular – are still unable to spend time building the necessary templates. While true document automation may be out of reach for these organisations, they can still use a kind of light version of document automation which we call “skeleton automation”.

With skeleton automation, you just select a high-quality precedent of a given type of document (e.g.: a Shareholders Agreement, an Outsourcing Agreement, etc.) and anonymise it. You can then choose to either eliminate the clauses from the document (figuratively preserving only the “skeleton” of the document) or keep the clauses in and regard them as standard clauses.

Do make sure that these clauses are not highly specific to the original precedent document, though. Some clauses may be compromises that come about through intense negotiations which neither party is happy with. Using these clauses as standard versions will risk pushback by the counterparty every time it is used, greatly extending the time to close.  

This skeleton document is then supported in the background by a collection of clauses that act as modular blocks that can be plugged into the document on an ad-hoc basis.

Where this collection of supporting clauses comes from depends again on how mature your organisation’s knowledge management expertise is and what your budget is for specialised tools.

There are three different options:

  • No budget, no expertise: the status quo in most legal teams. Lawyers dig into whatever precedents they can find to look for relevant clauses. Then they play the copy/paste/tweak game every time they draft a document. Having a skeleton document in this situation doesn’t really improve the process all that much, but at least it can provide a clean Word style to start from and a few useful base clauses. Over time, that skeleton may also grow to contain drafting notes from experienced lawyers.
  • Budget, but no expertise: a situation that many legal teams find themselves in currently. They have the money to buy new tools but not the time to invest in manually curating their knowledge. In this scenario, a clause extractor tool (see below) is the legal team’s best bet.  
  • Budget and expertise: a minority of legal teams – usually ones that have already passed the previous two options – will have experienced that significantly more value can be derived from curated knowledge management. Consider the legal team that wants to be able to distinguish between its standardized and fallback clauses. Or the Commercial department in a law firm that wants to consistently use the same clauses for the drafting work it does for one of its key clients. These teams will want to invest in a clause library tool (see below).

2. Document automation

Document automation technology helps lawyers get to a first draft in a fraction of the time it takes to do so manually. Depending on the quality of the automation and the negotiation power of the parties, this draft may even already be the final version.

Here’s an example of what a document automation tool can look like for the end-user.

Document automation in the style of Clause9

On the right-hand side, you have an interactive preview of what your document will look like when you export it to MS Word or PDF. On the left-hand side, you have a questionnaire that allows you to fill out information that then appears in the document on the right.

Caveat: the older generation of document automation tools is generally not able to show the interactive preview. Instead, users must first fill out the entire questionnaire before they can download the document. Then they have to review the document to see if it is to their liking. If it is not, they have to go back into the questionnaire and repeat the entire process. This naturally causes a significant dip in user-friendliness.

Underlying the interactive preview is a template document that can exhibit varying degrees of flexibility depending on:

  • the level of optionality built into the template (e.g. optional paragraphs, alternative clauses, placeholders, etc.), and
  • the sophistication of the document automation technology at play (e.g. repeating clauses, conditional logic, different language versions, etc.)

Never worry about styling again – Document automation allows you to produce perfectly styled documents every time.

More powerful document automation tools aren’t limited to only ensuring consistency in MS Word’s default styling settings. Some also allow you to customize and centrally adapt styling settings not native to Word. For example:

  • How numbers and sums of money are displayed – e.g.: “5.000 EUR” or “five thousand (5,000) USD”, etc.
  • How cross-references are displayed – e.g.: “Section 5” or “Article 5 (Liability)”, etc.
  • How dates are shown – e.g.: “5 February 2024” or “February 5th, 2024”, etc.

Time-consuming – Just as document automation saves time, it also consumes it. Someone needs to produce the base templates that power the automation. Someone needs to maintain them. Someone needs to learn how to use a document automation tool to perform the automation (although most legal teams tend to outsource this work to the vendor or an implementation partner).

This makes it very important to consider which documents merit the investment in automation. If you draft a certain type of document only a few times per year, and the manual process currently doesn’t take up that much time, then it will be difficult to recoup the investment if your only goal is to save time.  

Document drafting in the fast lane – It’s obvious that document automation can cut down on the time it takes to produce a first draft, but it is often underestimated how much time that is.

Some of our clients that use Clause9, ClauseBase’s advanced document automation tool, report that they produce documents up to 14 times faster. For context, that’s the difference between taking 2 hours to produce a draft, or taking only 8 minutes.

Complexity – As soon as you start venturing into more robust document automation tools, you’ll find that they become increasingly more complex to master.

Organisations that have successfully rolled out document automation technology have typically done so because they either have a dedicated team of legal engineers who have the necessary time to build expertise with an automation tool, or they’ve outsourced the automation to an expert.  

Rigidity & consistency– When producing a first draft via document automation technology, the only way to interact with the document that is being generated, is via the questionnaire. The automator puts a lot of thought into what that questionnaire needs to contain and what it doesn’t. That way, they map out the playing field for what can and cannot be done with the document.

This is perfect for in-house legal counsels who wish to outsource contract creation via a self-service platform since the business user cannot make any amendments that are not approved by Legal.

Rigidity & consistency – Document automation tools are deterministic in the sense that they can only leverage the exact options embedded in the questionnaire by the initial automator.

Suppose you’re filling out a questionnaire and in the interactive preview, you see a clause start with “The Supplier shall deliver […].”. You will not be able do those surgical text adjustments like “The Supplier shall use its best endeavours to deliver[…]” that lawyers are so fond of, until you’ve exported the document to Word or unless the automator has given you the option to change this in the questionnaire.

2.1. Tools for document automation

Document automation tools have been around for a few decades and a wide array of different tools for different use cases and different audiences has sprung up in that time.

As a result, the document automation landscape today is a spectrum, ranging from easy-to-learn tools that are limited in flexibility to virtually limitless tools that require days or even weeks to master.

Any investigation into different document automation tools should always be aligned with what your process currently looks like or what you are trying to upgrade your process to.

Are your templates limited in flexibility and are you primarily interested in creating a single source of truth where users can use the most up-to-date version of a template? In that case, a basic, easy-to-use tool may be your best bet.

Do you (want to) have well-developed templates with many different legal options available to the end-user with the purpose of achieving a first draft that is as close as possible to the final version? Then you’ll need to look at the more robust automation tools.

Do your needs centre around the “non-drafting” parts of the contracting process (e.g.: workflow automation, archiving, continuous management, etc.)? In that case, you need to consider whether you wish to adopt one or more “point solutions” (i.e.: focused tools that only do one part of the contracting process but do it really well) or “end-to-end solutions”, also known as “Contract Lifecycle Management” (CLM) tools (i.e.: jack-of-all-trades that cover all aspects of the contracting process but are typically weaker than respective point solutions).

Below, we look at different types of automation tools and concrete examples thereof in order of sophistication. Again, we want to emphasize that this overview is by no means exhaustive. If you want to explore additional tools, we recommend checking out the DocAutoDatabase.  

A. Form fillers

On the document automation entry level, we find form fillers. These are automation tools that do little more than offer a way for users to “fill in the blanks”. For example: the names and representatives of each party and placeholders like the commencement date or duration of an agreement. That does not mean they don’t have functionalities that enable conditional logic. These are usually just an afterthought, though, and not a main draw.

Due to the fact that form fillers offer limited automation flexibility, they usually only achieve real value (i.e.: beyond what a simple mail merge in MS Word can do) when combined with functionality that allow the forms in question to be filled out automatically – through an integration with a CRM database or as part of a wider Contract Lifecycle Management tool – or when the volume of documents that needs to be generated on an annual basis is sufficiently high.

That we emphasize the limited functionality in these tools in no way is meant to belittle them. They are intended for an audience that doesn’t need more than that and fulfil their purpose wonderfully. Imagine you are a legal counsel in a large organisation and you want to give your procurement team the ability to generate standardized NDAs where the only variables are the counterparty’s company information. A form filler is an easy and useful tool for that use case. They offer not just an efficiency boost but also the peace of mind for the legal team that consistency of the underlying template brings.  


Formstack is a workflow automation tool that offers multiple options for form creation, including static forms, surveys, and workflows, catering to diverse use cases.

Formstack markets itself to a broad range of different industries, showcasing that their platform is configurable enough to meet the needs of professionals whose way of working can often be vastly different.


Pandadoc places a strong emphasis on enabling sales teams and empowering the role other business units besides Legal have in the contracting process. As a CLM tool, Pandadoc offers document automation as only a single part of a broader offer that also includes e-signing, reporting, and post-signature contract management. Like most CLM tools, they are therefore much more than a just a tool for generating contracts but also primarily aimed at in-house legal teams as attorneys rarely have need of these additional functionalities.

Office & Dragons

Office & Dragons specialises in enabling the generation of multiple documents simultaneously, facilitating rapid customization and mass production of standardized documents. It is supremely good at this kind of bulk document generation of straightforward documents.

Office & Dragons can extract tables from placeholders which its AI recognizes in an uploaded Word document. Every column in this table is a new variable, every row a document for which this value can be filled out. You can also add columns manually if the AI doesn’t recognise a placeholder.

That said, it is clear that Office & Dragons focuses primarily on bulk generation in high quantity with limited changes and less on highly nuanced document creation for a single counterparty. It offers some level of conditionality, but this is primarily aimed at filling or not filling out a placeholder through an “if-then” interface of one level deep. Office & Dragons works very nicely with the styling present in a Word document, but of course that means the styling in this document needs to be completely on point. Not always a given in the legal sector.

B. Block editors

Most form fillers are not purpose-built for legal drafting. For lawyers who want to have the option to drop clauses in and out of a document or even generate suites of documents, these tools will simply not cut it.

Block editors offer a more robust experience for those lawyers. We call these tools “block editors” because they work by uploading a template document which then gets cut up into individual blocks that can be submitted to all kinds of conditions and optional logic.

The main value of block editors comes from their increased feature set and the fact that they are relatively easy to pick up. Block editors employ a “no-code” approach to document automation (as opposed to “low-code”, see below), meaning that you just use a point-and-click interface to set up conditional logic.

This lessens the steep learning curve for busy lawyers and works well as long as you don’t use too much conditional logic. If you do, you run the risk of having to implement all kinds of workarounds due to the fact that most block editors can’t handle different levels of conditional logic very well (for example: an optional clause, which contains an optional paragraph, which contains an optional sentence). Block editors will either force you to duplicate each possible variation of this clause or will allow you to embed it all in one clause, but make it very hard to untangle the web of conditions in the future.

Because they are so easy to learn, good CLM tools will offer them as part of their end-to-end contract offering that workflow automation, negotiation platforms, contract reviewing, etc. Nevertheless, lawyers with experience using block editors often report running into technological limitations fairly quickly.


Probably one of the most powerful block editor type document automation tools out there. Avvoka 1.0 launched in 2015 and the early version of the software struggled to deal with the typical technological limitations of block editors. A few years back, the company released Avvoka 2.0 which provided a significant boost to the platform’s functionality.

Avvoka allows you to upload a template to its platform and identify all the different optional, variable, and conditional elements using a simple no-code interface.

While not a true CLM tool, Avvoka also offers features that go beyond the creation of an initial draft document, which is where most document automation tools stop. As such, it also allows parties to negotiate a document on its platform and offers data analytics on documents generated with it.

Avvoka manages to strike a complex balance between power and ease-of-use – lawyers want to tailor a document down to its details, but they also want to not be inundated with different buttons and features. A lot of the more advanced features on the document automation spectrum are also there, such as repeating clauses, secondary conditions, API integration and centralised house styling.

Thanks to its no-code approach, a great deal can be achieved through a simple point-and-click interface. This point-and-click interface does have some weaknesses, though. Most notably, it greatly complicates or even limits interactions within a clause. For example:

  • Say you want to give a user the option to choose how “the Supplier” should be called in a document (i.e.: as “the Supplier”, “the Provider”, by name, etc.). With Avvoka, you would have to manually create a condition for each individual reference to this term, complete with conditional logic explaining how the word “the” should disappear if the party is referenced by name and how “the” should be capitalised at the start of a sentence.  
  • Avvoka is language-agnostic. This is another way of saying that it doesn’t understand language. The benefit is that you can automate documents in any language. The downside is that there is no inherent linguistic support built-in. Suppose you want to change “the Seller” to its plural version “the Sellers”. A lot of different verb and noun conjugations will need to be manually “programmed” into a document to this. And that’s just for English. If you work in gendered languages like French or case-based languages like German, this is an absolute nightmare to configure manually.  
  • Mathematical calculations and advanced conditional manipulations are inherently trickier in Avvoka. Asking the tool things like “if the date filled out by the end-user is on a weekend or a bank holiday, then it should be changed to the next business day” is not possible.

Juro is a comprehensive CLM tool that caters to various sectors, including legal, HR, sales, procurement, and finance. Like other CLM tools we have covered in this guide, Juro too focuses on more than just contract generation. It covers the entire contract lifecycle of intake, creation, signature, and storage.

There are many CLM tools that offer document automation in the style of a block editor. Tools like ContractPodAI, LinkSquares, and Ironclad have stormed to the top of the CLM market in recent years and offer a similar approach. As a result, while there are some tools that do one thing better than another, they are pretty comparable on the feature set. We’re using Juro as a benchmark here, but many of the comments can be extrapolated to other modern CLM tools.  

Since CLMs focus on high-volume contracting for in-house legal professionals, their concern is less with offering a robust feature set that can handle any kind of possible legal nuance and more on creating an easily configurable setup experience. As a result, Juro does not offer the kind of flexibility that a more powerful block editor like Avvoka does.

Nevertheless, in-house lawyers will be pleased to hear that basic conditional logic can be easily configured. The major downside of Juro’s block-editor approach is that conditionality multiple levels deep (e.g.: an optional clause that contains an optional sentence) requires some workarounds and “hacky” solutions that involve duplicating clauses. That works up to a certain degree, but quickly becomes incredibly cluttered.


Lawlift is a flexible no-code block editor launched in 2015.

Lawlift is on par with Avvoka in terms of the level of flexibility it offers for automators (notably, it launched in the same year). That said, it opted to be slightly more on the easy/simple end of the document automation spectrum. As a result, it offers a sleeker, more minimalistic interface, but less automation power when compared to Avvoka.  

For example, Lawlift does not offer:

  • dynamic table configuration (e.g.: creating a list of shareholders where you don’t know up front how many rows for how many shareholders there will need to be),
  • batch document generation (e.g.: generating 50 employment agreements for 50 different employees in one go),  
  • internal database creation (e.g.: automatically filling out the signatories to a contract based on a database/spreadsheet stored within the tool itself; this can only be achieved in Lawlift via a PowerBI integration).

On the other hand, as a German company, Lawlift is much more sensitive to the linguistic needs of non-US/UK client. As a result, it features bilingual drafting support which is superior to most other block editor automation tools (which typically avoid the topic of linguistic support by sticking to being “language agnostic”).

That said, this language support does little more than allowing different language versions of the same document to be exported in two columns. This is still inferior to tools that provide deep, grammatical support (e.g.: automatically conjugating verbs, (pro)nouns, etc. as necessary without any conditional logic programming).

C. Low-code template editors

The original legal document automation tool that set the tone for all others to follow was Contract Express and it has been around in its earliest form since 1996. It was Contract Express that solidified the blueprint for document automation in the typical setup of a questionnaire that interacted with a flexible template.  

It did so in the form of what we call a low-code template editor. It starts from a Word document that acts as the template underlying the questionnaire. Users can add variability to it in the form of placeholders, conditions, and other types of logic by using the designated “syntax” of the tool – i.e. the different symbols and formulas used to clarify this flexibility. This is why they are called “low-code”. It’s not quite programming, although it may definitely feel like it to lawyers who have never had an introductory programming course.

Most low-code template editors present a significant leap in functionality compared to most block editorsbecause they offer a much more granular way to interact with the text of a clause.

That said, they also feature a notoriously higher learning curve because you need to learn the syntax. Many of the block editors outlined above came about as a response to this weakness and to offer a more user-friendly, albeit more limited, alternative. As long as templates must only be superficially modified, block editors are a good choice. But as soon as texts need to dynamically changed, there is no way around low-coding tools.

Contract Express

Contract Express is the original document automation tool for the legal sector that the modern crop of document automation tools all looked to for inspiration.

You could conceivably give this title to another popular document automation tool called “HotDocs” as well. They even entered the market before Contract Express did. Ultimately, though, they haven’t had the same impact in the legal sector as Contract Express.

Today, many of the largest law firms in the world still have a Contract Express license and dedicated teams of legal engineers specialised in automation via Contract Express. That doesn’t mean it is actually used across their entire office, but we have met fairly little law firms in that upper segment that didn’t at least have a first experience with document automation. For a majority of those firms, Contract Express presented that experience.

With its low-code approach and vast feature set that it has built up over decades, Contract Express is primarily suited for the more sophisticated type of document. It allows for far greater textual control within paragraphs compared to the average block editor. This translates to such things as automatic calculations, conditions within conditions and central management of your data fields.

Since Contract Express offers users a range of advanced features and allows them to apply these features in a highly granular fashion, it unlocks an entirely new dimension of flexibility compared to the average block editor. This is the blessing of Contract Express’s low code approach...

… but it is also its curse. As you can see in the screenshot above, there is a distinct “programming feel” to the syntax of low-code template editors like Contract Express.  

Despite how similar the programming mindset is to the legal mindset (in all its variations on “if X then Y”), most lawyers have no desire, skill, or time for learning how to use a low-code template editor’s syntax.

As a result, the larger international law firms, where Contract Express has thrived more than anywhere else, employ entire teams of dedicated legal engineers whose primary task is automating and maintaining these templates. This is a concern that is shared across all low-code template editors. Hence the rise of the no-code movement in recent years.

Nevertheless, there are also several issues unique to Contract Express which it, despite its years of head start, hasn’t managed to do very well:

  • Interactive preview – Users report that the Contract Express interactive preview is slow to load. This means that every time you field out a basic field like a party’s name or address, the tool has to reload the document. Not a dealbreaker at first, but if you have to wait 2 seconds after each question in a 100-question interview, it starts to grate.
  • Conditions management – Contract Express is one of the more flexible document automation tools out there, but they don’t offer enough to make life easier for automators who are operating at the far end of that flexibility spectrum. Having used Contract Express ourselves when we were still lawyers, we are reminded of one situation where we automated a form for which the end result could only ever be 2-3 pages of text. The template that used the Contract Express syntax to power the automation was over 1.000 pages (!) long, because all the different variations need to be written out in full with a lot of repetition along the way.  
  • Flexible styling – a document’s style in Contract Express is intrinsically tied to the style as set up in the marked up Word document. This is a missed opportunity to introduce centralized, flexible styling as modern document automation tools like Avvoka and Clause9 have done. This allows you to set up styling centrally and deploy them to the same document in different situations. For example: suppose you are an in-house counsel working for a group of companies. Sometimes your Master Services Agreement needs the house style of Company A; sometimes that of Company B. You want to be able to have the right style with a click of a button. That’s not possible in Contract Express.

Even more disconcerting however, given its status as the ubiquitous document automation tool, is the fact that Contract Express is built on ancient Windows technology, using the old MS Word plugin architecture, which is slowly dying out.

In Contract Express, the preparation of the template is done entirely in Word with the help of a plugin of the old generation of MS Word plugins. These old plugins (called “COM” and “VSTO”) work by spreading their tentacles deep into the bowels of MS Word. This allows them to manipulate MS Word in almost any way they want, in the same way a puppeteer manipulates his puppet’s strings.  

The plugin's freedom to do whatever it wants, has significant drawbacks:  

  • No interactive testing: to test a template, it must first be loaded into a separate browser environment. This back-and-forth between Word and browser causes significant slowdowns.
  • Instability: any programming mistake can cause Word to crash, or cause mysterious errors.  
  • Clashing instructions: Plugins cannot play together nicely, and can easily step on each other's toes. Again, causing crashes.
  • Complex Updates: Patching these add-ins isn't straightforward. IT admins often struggle with compatibility issues.

The new plugin architecture features none of these problems, which is why Microsoft is pushing them so hard. But just because this new technology exists, does not mean that vendors are spending time and money converting their existing applications towards the new plugin architecture. In fact, for many existing plugins that have been on the market for quite some years, this conversion is simply impossible to do, because a lot of functionality they rely on is deliberately forbidden territory in the new plugin architecture.

This is the reason why many vendors remain deliberately silent on this topic. In any case, Microsoft tends to support legacy technology for some years, so they still have time. Microsoft will pull the plug eventually, though, (it has already done so for the old Outlook plugins) and it’s anyone’s guess how Contract Express will deal with this when the time comes.

Finally, it pains us to say this, but — even leaving aside its technical age — the future of Contract Express seems unclear. The product hasn't seen any major updates in years, and we often hear users complain about very poor support. The product team has also undergone several moves within the company, so it seems to us that Thomson Reuters just doesn't know where to go with this product. It would be very said that this once so innovative product would die a slow death, being positioned as a run-off-the-mill drafting editor for Thomson Reuter's new favourite child, HighQ. 


ClauseBuddy is part of the ClauseBase suite of legal drafting products. It is a Microsoft Word plugin built on the new Office plugin architecture that functions as an all-in-one toolbox for legal drafting.

One of those tools in its toolbox is the “Smart Template” module, which allows users to automate documents directly from within MS Word.

To indicate flexibility in a ClauseBuddy template, users simply have to highlight their data fields in blue using MS Word’s native highlighting functionality. These blue highlights then inform ClauseBuddy that certain information needs to be filled out there. For more advanced kinds of flexibility, ClauseBuddy uses basic syntax rules to visualize the typical “if X then Y” type of conditional logic.

To understand ClauseBuddy’s position on the document automation spectrum, it’s probably good to know that it’s a mix of a simple no-code approach, sprinkled with advanced features of ClauseBase’s advanced document automation tool “Clause9”, which is situated on the powerful/complex side of the spectrum.

When rolling out Clause9 to lawyers, we realised that many of them hadn’t developed flexible templates yet – a prerequisite for advanced document automation technology. Even among those teams that had developed these templates, they always had a number of documents that didn’t require a lot of conditionality, optional annexes, or flexible styling, etc. We also noticed, however, that as lawyers spend more time developing their templates and becoming accustomed to document automation, their desire for additional flexibility skyrockets.

Our aim with ClauseBuddy was therefore to offer an upgrade path. A way to start off with limited, simple automation but with room to grow to Clause9 as templates are developed further.

As a result, ClauseBuddy’s highlight-based markup excels at very quickly automating basic placeholders. Creating conditional logic is a bit trickier, as users then also need to learn its rudimentary syntax and how to apply it. This also means that there are kinds of flexibility that are impossible to add to a ClauseBuddy-automated document. Some examples include:

  • Centralized house styling – as opposed to centrally defining styling rules and pushing them out to each document automated in the platform, ClauseBuddy relies on the user to ensure that each individual document has received a clean style in MS Word before automation begins; unfortunately, not many lawyers are capable of this)
  • Dynamic tables – this means the possibility to have columns, rows, and cells be able to drop in and out of a document based on conditional logic
  • Suites – this relates to generating multiple documents through one questionnaire (e.g.: a main agreement with a number of optional annexes)  

Gavel (previously known as “Documate” and before that “HelpSelf Legal”) is a low-code template editor with a no-code interface. This may sound confusing, but is quite logical if you look at how their template automation works (see this collection of GIFs on their site).

In short, you build your template in MS Word through a Word add-in (the no-code interface) which then includes the code in the Word document for you (which then becomes a low-code template). This is done by simply selecting text (e.g.: the name of a party) and indicating in the variables menu what kind of variable you want to create.

To a certain extent, this creates a “best of both worlds” experience. Creating simple placeholders with Gavel is incredibly easy and doesn’t really require you to understand the “code” the software is putting into your document. Basing the exercise in MS Word, home turf for any lawyer, also makes it easier to get started.

That said, as soon as thing start getting a little bit more complex or as soon as you want to tailor your document beyond simple placeholders and basic conditional logic, you will need to get acquainted with the code. For example: when creating conditional tables. Furthermore, Gavel does not feature the option to have centralized styling, meaning that your automator needs to be really good at dealing with MS Word styles and make sure that they are automating a clean document, or issues will inevitably arise with numbering and formatting.

Gavel has been around since 2018 which means that it had time and opportunity to not only develop a robust product that can handle the more advanced text manipulations of the document automation landscape (e.g.: repeating lists, dynamic tables, nested conditions) but also expand its scope beyond just automating documents. For example: building on its history as a tool that offered legal apps for low-moderate income consumers, Gavel has differentiated itself from many other document automation tools by focusing on giving law firms the means to create legal products in the form of client portals.

Gavel has two major weaknesses, though:

  • Similar to Contract Express, automation is done through the combination of a Word add-in (where a template is prepared) and a browser-based platform (where the automation is done). This means that automators constantly have to switch between different contexts when they are automating, testing, and maintaining their documents. Users that we have spoken to indicate that this makes it difficult to troubleshoot and can make automation far more mentally taxing because they have to remember how they set up the automation in one part of the workflow when they switch to another.
  • Surprisingly, despite being of the newer generation of automation tools, Gavel does not feature an interactive preview of the template when filling out a questionnaire. Not a problem for basic forms, but this has a few serious downsides for more flexible documents.

As an automator, you need to go through the entire list of questions to ensure you have a functioning document. When testing, this means that you sometimes have to export a document 40-50 times with different answers to different questions to make sure there are no minor glitches or errors that snuck into the automation.

As a user, this means that you have no idea what the document will look like until you hit the export button. Could be you made a mistake and don’t catch it until you have completely finished filling out the list of questions. Could be there is something wrong with the automation and – again – you’ll only know after having spent 5-30 minutes filling out a questionnaire.

D. Clause-based automation tools

The most advanced tools on the document automation spectrum use a “clause-based” approach to automation, rather than the “template-based” approach as adopted by the previously mentioned categories.

Rather than taking a static, monolithic template document and then adding flexibility to it using low-code syntax or a no-code point-and-click tool, clause-based automation tools instead treat templates like stacks of individual building blocks. You start by uploading clauses to the platform which you collect in a central clause library. The templates you wish to automate are then (re)constructed by stacking these building blocks on top of each other and adding the necessary flexibility, either via low code or no code.  

The clause library sits at the heart of the automation platform. One thing that makes clause-based automation tools so powerful is the inherent flexibility that comes from the “plug-and-play” clause assembly that comes with this modular approach.

But perhaps the most powerful differentiator is that the clauses can be reused centrally across different template documents. This means that if you need to perform an update to a clause that is used across multiple templates, you only need to perform that update once. The change will then ripple through to all documents that make use of the clause. This also provides a compounding benefit to your automation efforts: for each template you automate, your clause library grows, which can then make the automation of new templates easier because you already have the building blocks right there in your library.

Finally, this collection of standardised clauses can then also support other parts of the legal drafting workflow, making document automation technology work for the lawyer beyond the first draft.  

If standardization and central clause management are key to your automation plans, then clause-based automation tools are the way to go.

Do note that some tools in the other categories also allow users to create clause libraries. Block editors like Avvoka and _Ment, for example, added this feature as part of the continuous improvement of their tools. The difference with true clause-based automation tools, however, is that this modularity is built as an afterthought. The default way of working is still to first upload your monolithic template and add conditionality there. Then, optionally, remove blocks from them and replace them with blocks from the clause library. This application of clause libraries is more limited in scope: it will not provide the flexibility of a fully modular tool and is not available directly inside MS Word to support with drafting on third-party documents or after the first draft has been created.  


Clause9 is ClauseBase’s original product with which the company entered the legal tech market in 2018. It is a low-code, clause-based automation tool that differentiates itself on its extremely high level of flexibility, its modular approach to automation via a centralized clause library, and its advanced API integration capabilities.

It works by uploading Word documents to a web-based platform and then importing select clauses from those documents for inclusion in a central clause library. This turns the clauses into intelligent building blocks capable of flexible styling, flexible numbering, flexible terminology, and more. Templates are then first created by stacking the right building blocks on top of each other and assigning conditional logic. These can then be transformed into a questionnaire in the style of traditional document automation.

Clause9 furthermore enjoys all the benefits of clause-based automation tools as outlined above, most notably in the management of templates through these centralized building blocks. An update to one building block used in multiple documents automatically ripples through all these documents (in all languages those documents support) to ensure changes stay consistent.

When we started building Clause9, our goal was to improve on traditional document automation tools, most notably Contract Express, after a lacklustre experience using them when we were still lawyers. We felt that a modular approach with centralized building blocks and centralized styling aligned much better with the way that lawyers wanted to do document automation. More importantly though, we felt limited by traditional document automation tools. Lawyers want to tweak every letter, every comma, and every minute detail of a document’s style. We were no different. We wanted something that allowed us to customize the document much further than what even Contract Express, the golden standard at the time, could do.  

We’re happy to say that we achieved our goal of creating the most advanced document automation tool on the market. But of course there are downsides to this focus on advanced drafting. 

Looking back at the spectrum of document automation tools identified at the start of this chapter, you’ll remember that the more flexibility a document automation tool offers to its end user, the more difficult it becomes to master. Clause9 is no exception. Just as it offers lawyers untold flexibility in their automation efforts, so it also offers one of the more steeper learning curves in legal tech to its power users, roughly similar to the level of Contract Express.

This is mostly due to the expansive nature of Clause9’s syntax. While its main principles are captured in a single one-pager, the options attached to those principles have grown tremendously over the past years due to client demand. Take Clause9’s “special functions” which allows text to be manipulated in hundreds of different ways – from turning a list of items into bullets, to calculating whether a specific date falls on a weekday, to forcing terms to be shown in their plural form, and much, much more. In 2018 we started with a handful of special functions. Now, this list has grown to the size where no single Clause9 user knows all available special functions by heart; relying instead on the central special functions resource.

Clause9 is therefore most enjoyed by dedicated automators like legal engineers, people with prior experience in document automation, and legally trained people who have had an initial exposure to principles of programming.

For those organisations interested in getting started with document automation who don’t have any of the former, we typically refer to ClauseBuddy’s Smart Templates functionality, which is a mix of a no-code tool and a light version of Clause9’s automation engine designed to sit at the simpler end of the document automation spectrum (see above).


WeAgree is a CLM tool which originally started as a document automation tool and has been around for over 2 decades. In that time, it has grown tremendously in scope, encompassing not just the creation of documents, but negotiation, workflows, archiving, storage, reporting, e-signing, and more.

In the typical style of clause-based automation tools, WeAgree treats template documents as stacks of intelligent building blocks and allows for template building through a centralized clause library.

WeAgree takes some different design decisions than Clause9, the other document automation tool in the clause-based category. Most notable is its no-code approach where Clause9 opts for low-code. Given its target audience of primarily in-house legal teams, this makes sense. The increased flexibility offered by low-code automation may be less interesting in environments where a lot of drafting language can be standardized and focus is on managing volume instead of legal variability. In that situation, the ease-of-use of a no-code tool is preferable.  

Perhaps one of the biggest weaknesses of WeAgree is its graphical user interface. When ClauseBase launched Clause9 in 2018, we believed we were the first document automation company to create clause-based automation. Little did we know that WeAgree had us beat by 12 years.

This is a testament to the vision of WeAgree’s founder (a former lawyer, unsurprisingly). But it is also where the cause of its biggest weakness becomes apparent – a clunky, and outdated interface.

The entire WeAgree interface clearly dates from the early 2000’s and has undergone little modernization since then. In fact, the feature expansion that the product has seen since then has only served to make the interface more crowded and make it harder on the user to learn how to use the tool.

WeAgree seems to understand this – having created hours of video tutorial material. The question has to be asked, however, if lawyers are willing to spend these hours learning how to use the tool. Given WeAgree’s relatively limited market penetration in the 15+ years it has been active, we are inclined to think not.