SaaS companies should use external conversion optimization systems

This is a concept that always reminded me of a very traditional, bricks-and-mortar style accountancy (or insert alternative ‘traditional’ industry here) firm of the early-to-mid-2000s - that have recently decided to move away from a paper-heavy and labour-intensive administrative approach - not something I would have expected to continue as a norm for many SaaS companies in 2018, when one considers the explosion in availability of SaaS in the past decade or so. However, you will find that the existence of these purpose-built, in-house systems in modern SaaS companies are still very much commonplace. When we now live in a world where there is a SaaS solution for any business pain point, it seems like somewhat futile to exert time and effort into constructing a system, characterised by a clunky and unintuitive UI/UX, frequent bugs and pervaded by sluggishness. It is understandable that, in building such a system, there is the primary benefit of running a system that is tailored exactly to a companies needs (assuming that the system is fit for purpose). However, I have found that, the benefits of the in-house system tend to stop there. Every other factor associated with in-house systems seems to be negative. Below I have examined some of the drawbacks associated with a continued dependence on building and operating such systems in the present day.

Software developers' time has to be one of the greatest opportunity costs of building in-house models. Software developer time is at a premium due to the scarcity of personnel possessing the necessary skills. The initial build time of a proposed system could take months. That is months lost that could alternatively be spent on optimizing the companies core software product offering, or augmenting the product with new features. I’ll try to illustrate the negative effects of this with an example:

Imagine a hypothetical B2B SaaS company has decided to build an in-house model to record prospects progress through the sales pipeline. It is somewhat detailed and feature-rich, so this takes three months to complete. After the initial build, the sales agents start to implement the model into their daily activities. After a week of testing, the sales agents find that the initial product doesn’t fit well with their existing workflow, and changes are necessary. Back to the drawing board, for a while at least. Another few weeks of changes, based on the agents feedback, the product is ready again.

The sales agents start to use the system. The product is now more compatible with their existing workflow, but now there are a number of bugs that are occurring, due to the changes implemented in the original system. The developers don’t have time to address the bugs immediately, so the agents have to persevere for a week, and try to work around the bugs - this is possible but it severely slows their daily progress and has cost them a significant amount of productivity for the whole week. The week after, the developers have time to fix the bugs that arose with the changes to the original in-house system. Included in the feedback from the sales agents is a feature request for a basic user conversion predictor to filter high probability users and low probability user of conversion during the trial period. The VP of Sales agrees with her agents that this is imperative addition to the model. The data analysts in the Business Intelligence team now work in parallel with the software developers (who are fixing the bugs) to develop this model, which delays their progress on an important competitor analysis report. Both the teams complete their bug fixing and prediction model work at approximately the same time. The system works perfectly for two weeks without issue.

On the third week, the SaaS company in question launched some major feature changes to their core offering and, in addition, launched an entirely new product to complement their existing products. The in-house sales pipeline system has been specialized for the existing offerings - it does not accommodate for the new product or acknowledge the feature changes in the existing product. More work for the software developers, that could be better spent on the core product suite. After two weeks of observing the user conversion predictor model, the agents agree that the model is inaccurate and thus completely ineffective. Therefore, in addition to the piling workload of the software team, the two weeks of effort by the Business Intelligence team was effectively wasted, and their important core work on the competitor analysis report has now been significantly delayed.

As you can see, there will be a never ending cycle of new implementation and ongoing maintenance of the in-house model. Had the SaaS company considered one of the many 3rd party platforms for recording prospect progress through the sales pipeline, there would be a relatively small financial outlay per sales agent (perhaps at worst, $30 per agent per month). However, the developers would have been able to devote more time to the main product suite, and its ongoing maintenance. Also, in the smaller chance the team ran into issues with bugs with a 3rd party platform, it is likely that the platform provider would have responded quickly to fix the bugs in the interest of high quality customer service, as the SaaS company would be a paying customer of the platform. In addition, whilst the firm way have had to slightly change their workflow to fit the UI/UX of the 3rd party platform, it wouldn’t be a major shift, as these platforms are designed to fit the workflow of thousands of firms in a similar space.

One can see how this opportunity cost of valuable, specialized employee time has affected the software developer team, spilled into other departments, namely the business intelligence team. In addition to this, the sales team’s productivity was worsened during the period whereby the system encountered problems, and this would likely have an impact on the firm's revenue. Another factor that wasn’t covered in the above example, was the firm’s ethos when requiring such tools, that is, if the sales team had an in-house tool developed, it is likely that the firm would build everything in-house for all departments: perhaps a helpdesk tool for the support agents, an accounting system for the finance department - where does the impingement on valuable company resources end?

Whilst Traitly do not offer in-house solutions for all business pain points, we do offer a powerful AI conversion prediction model - allowing you to see what users are most likely to convert, based on what successful users have done in the past. It will also determine what next key actions are most important for user conversion, so they can be intelligently targeted with personalized messages via your existing tools such as Intercom, HubSpot, Facebook and Twitter.

Subscribe below to get more useful information on conversion and retention.

James Moran

James Moran

James is the Customer Success Lead at Traitly. Previously, he worked at xSellco, a fast-growing SaaS company. Given his engineering background, he takes analytical approaches toward customer success.

Read More
SaaS companies should use external conversion optimization systems
Share this

Subscribe to User Onboarding - Conversion Optimization | Traitly