Today, off-the-shelf solutions for CX – many of which you can buy on a subscription basis – are more powerful than ever before.

Understanding what your customers think and feel is what makes you a truly customer-centric organisation. Businesses that understand their customers’ feedback make customer-centric decisions. And the best SaaS CX technology (like Chattermill) enables that.

Using artificial intelligence (AI) and Natural Language Processing (NLP), the pinnacle of CX software can analyse all your customer feedback touchpoints in real-time, providing in-depth analytics insights that can enable you to make your customers happy and your teams more efficient. This makes them a simple, accessible, and compelling option for leaders who want to improve their business’ CX.

At the same time, if you work for a high-growth business or an established enterprise brand (as many of you do), you’ve likely got a team of machine learning, NLP, or AI specialists that could potentially build a custom solution in-house. And when your business is paying out thousands every month on their salaries, there’s likely pressure from above to use that resource instead.

So what’s the answer for CX leaders in this mental bind?

In this article, we’ll discuss the arguments for and against buying and building CX software. And hopefully, by the end, you’ll be able to untangle the complex threads and come to a concrete conclusion. Right, let’s begin...

The Argument for/against Building Your Own CX Software

The argument for building in-house

There are a couple of reasons why you’d choose to build your own CX software in-house.

First, if you’re a big company with readily available data science and engineering expertise, you already have the technical capabilities to build what you need, without needing to look externally.

You may also have some pressure from above to use this resource, rather than looking to an external tool. After all, if you’re paying for expertise, why not use it?

Second, building a tool in-house does mean you can build and customise it to a specification that aligns with your CX needs at the time of production. This is particularly useful when you’re building a product that’s highly business-specific, or so niche that a SaaS solution isn’t available that can solve your problems.

Finally, building a proof of concept product can, in the short term at least, be less expensive than committing to a long-term contract with a SaaS tool.

The wedding scenario – the reality for most businesses

Imagine you’re planning on going to your best friend’s wedding.

You’ve visualised the outfit you want to wear. You know what materials you want it to be made with. You know what cut you want the trousers to be. You know your exact measurements after several unfortunate incidents with an inflexible tape measure. And your soon-to-be-wed friend has some handy sewing skills that you could make good use of.

It’s the perfect scenario to make a highly customised outfit that you know is going to be perfect, right? Well, perhaps not. Here’s why.

First, your soon-to-be-wed friend is likely to be quite busy. Potentially, just by asking you run the risk of deep resentment and a rescinded wedding invitation.

Second, your perfect wedding outfit will likely only be perfect for this one occasion, and won’t fit in a few years’ time. Some might say it’ll be unusable – embarrassing even.

Third, it’s going to be expensive. Making everything from scratch takes time and effort, and custom materials don’t come cheap. Inevitably, you’ll end up having to make some compromises, too.

There are probably many other factors we’re missing, too.

If you’ll excuse the extended metaphor, the same issues often crop up when building CX software in-house.

Not all businesses have the resources to build an effective CX tool

As discussed above, there’s no doubt some businesses have the technical capabilities to build a bespoke solution that aligns perfectly with their tech stack and business model in-house. But many others don’t have the resources to properly pull it off.

Machine learning, NLP, or AI specialists in most businesses are pulled in many different directions, with almost every business department vying for their time and technical know-how.

The result is it can take months or even years to get a CX solution on the development roadmap. Hiring more is expensive, too. And by the point it’s on the development roadmap, the functionalities and features you need to improve your CX will most likely have changed.

From experience, some features will also likely need to be prioritised over others due to time and resource constraints, which means it’s unlikely to get everything you requested.

“The real challenges of building in-house lie beyond the proof of concept. Data scientists are often good at building a prototype, but they are not specialists in building usable products. Taking it from a proof of concept to a fully functioning piece of software usually requires engineering and product expertise, and it is rare that in-house teams can easily pull teams together to build a working CX product. Not only because it’s difficult to access their time, but because in the long-term, it massively increases costs.” - Aji Ghose, VP of Data & Research, Chattermill

Adding features eats up time and money, if they happen at all

In theory, custom-built software can allow you to add or remove features as you grow. And they can be highly personalised to your own business.

But in many cases, once the software is built, the team disbands to work on other projects. And even when that doesn’t happen, every improvement to the software uses up additional time, resources, and money. Factors that are rarely considered from the outset.

There’s also the time it takes to actually build and implement the custom-built CX product in the first place. Though proof of concepts can be built relatively quickly, even with highly agile, time-rich resources, building CX software that’s usable can take months or even years – especially if you’re building in artificial intelligence and neural networks.

AI and NLP are changing fast, and it costs to keep up

Natural language processing is changing incredibly quickly. If you built an in-house piece of CX software using NLP a year ago, it’s likely now redundant.

The advantage of an NLP CX tool like Chattermill is the technology is modularised. It allows us to take things apart, iterate, and add additional features exceptionally fast. Take, for example, our recently announced insightful filter, and popular workflows.

In-house solutions are often stuck at the point they were created, and infrastructure costs increase massively as time progresses. And if you were trying to build a modularised system, costs would be high from the outset because you’d need to build a real engineering solution.

It’s also worth noting that if you’re setting out to build a machine learning system for CX, it may be more complex than you think. As Google argues in their Hidden Technical Debt in Machine Learning Systems paper, in any AI system, data collection, cleaning, verification, and infrastructure all need to be in place to support whatever machine learning you want to do.

In other words, it’s no simple matter.

Building an in-house tool costs more than you think

Finally, and perhaps most importantly, there are the costs.

In the mid-to-long term, custom-built software is more expensive than buying a SaaS product. And if you build in-house, you’ll have to pay to maintain it, too. According to the Standish group, 53% of projects cost 189% of the original estimate. This is something you won’t have to worry about with a SaaS solution.

A quick summary

If you’ve got a big budget, with machine learning, NLP, or AI specialists on hand to work on whatever you ask of them, and with a flexible timeframe, building a custom solution may be an option. But it remains complex.

For most CX leaders, building a CX tool in-house is unlikely to be a realistic possibility.

The Argument for Buying a SaaS CX Solution

Let’s start with the biggest consideration for most companies: cost.

Buying a SaaS solution costs much less in the mid to long term

Most SaaS CX solutions require little upfront investment. Indeed, most (like Chattermill) come with free trials. This will be advantageous for most businesses.

With a SaaS solution, you can test the tool out before committing budget. Even beyond the free trial, SaaS solutions cost less than they would to build in-house – in the mid-to-long term at least. Plus, the fact you can often pay monthly means it’ll have a lighter short-term impact on your CX budget.

Why are SaaS tools cheaper? Ultimately, it comes down to economies of scale. SaaS companies have thousands of customers, which means they can charge end users a lower price. This means you ultimately get more for your money. And, if you choose the right product, it’ll get progressively better value as the product improves over time.

Maintaining a SaaS solution costs you nothing

A second point to make about SaaS CX solutions is maintenance. If you build your own CX tool in-house, you’ll have to shoulder maintenance costs. These are often more expensive than you might expect – especially if you’re building a tool that utilises AI.

With a SaaS solution, however, maintenance costs are included in the package, which means that you can focus on what matters most – bringing value to your customers.

The control misconception

A common doubt around buying a SaaS solution is control.

It’s true that with some SaaS products, you have less control over feature improvements and product development. And it goes without saying that you have full control when you build in-house.

However, if you choose the right SaaS platform, one that treats their customers as partners rather than out-of-sight users, you’re able to have an open dialogue on what improvements need to be made.

It is, after all, in the interests of the SaaS company to improve its product in line with its customers’ needs. And you’ll also get the shared benefit of product improvements driven by the SaaS tools’ other customers.

Want to get results fast? A SaaS solution is probably the best option

A further case for buying rather than building a CX solution is speed of implementation.

Building a fully formed AI-powered CX technology in-house can take a long time. We’re talking years, rather than months. With a SaaS solution, however, you can be up and running within a matter of weeks. Of course, it takes some time to get the absolute most out of it, but if time is of the essence, a SaaS solution makes a lot more sense.

Want ongoing support? Buying a SaaS solution wins hands-down

Finally, let’s consider support. One of the key advantages of committing to a SaaS solution is the support you receive from dedicated customer success managers that know their software inside-out. So when you have a question about how to make the best use of the tool, or even want advice on your CX strategy, you have an expert close by. This can be invaluable.

When you build in-house, unless you have a team dedicated entirely to your custom-built CX software, this is unlikely to ever be the case. Even with the best organisational structure.

Conclusion

So should you buy or build CX software?

Of course, we’re a little biased. We see every day the value Chattermill brings to our customers, we delight when their businesses improve their CX, and when our users become personally successful off the back of it. This drive to keep delighting our customers everyday, helping them tangibly improve their CX so they can see the direct impact of growth and customer loyalty, is hard to replicate.

But even if we look at the build vs buy question objectively, for most businesses, buying a SaaS solution is the right thing to do.

To sum up:

• Building CX software in-house is more expensive in the mid-to-long term
• Building CX software uses up a lot of internal resources that you might not have
• Maintaining CX software built in-house can be difficult and expensive
• Buying a SaaS solution gets you over these hurdles, improves over time, and comes with ongoing support!

Want to learn more about Chattermill?

If you’re reading this, you’re more than likely looking for a CX solution to your big CX challenges.

One of the biggest challenges right now for CX leaders is making sense of millions of pieces of customer feedback data. And here’s where Chattermill can help.

Strap yourself in for a sales pitch.

At Chattermill, we help our customers unify, measure, and analyse all their customer feedback touchpoints in real-time, providing in-depth insights that enable you to make your customers happy and your teams more efficient.

Don’t take our word for it. We’re used every day by the world’s fastest-growing, customer-centric brands, including Wise, Uber, HelloFresh, and Just Eat. And we all know how good these brands are at CX!

Want to see Chattermill for yourself?

Sign up for a free trial here.

Or if you’d rather speak to a real person, book a personalised 30-minute demo and our CX experts will take you for a spin!