My Dynamics 365 Business Central priorities for Microsoft in 2020

January 26 2020

At the start of any new year, most of us set targets for what we want to achieve by the end of it. This year more than most though, I've been struck by how, as a predominantly Dynamics 365 Business Central partner now heavily SaaS-focused, we are dependent on Microsoft more than ever if we are to achieve our potential.

That has led me to compose a list of 'must-haves' for Microsoft that I thought I'd share. These are not specific wishes (I would put those on the BCIdeas site for voting on) but more big-picture-strategic approaches. It would be great if it starts a discussion. Am I way off beam? What's evident to you that I've missed?

1. Give us control of application updates

Users don't know or care about the platform if it runs. What they do notice is when the application changes, especially when it doesn't do what they expected. Microsoft made a significant advance last October by moving the application into an extension. What we need now is for partners or end-users to be in control of when to upgrade that base application extension rather than Microsoft.

That Microsoft automatically applies its cumulative updates to the platform every month is fine, provided of course that they don't break the application. It's then fact that they also automatically apply the application updates as well, and you have no idea it's going to happen until you get the email saying it failed or it’s done.

What we need is for the client or partner to be able to choose when to apply the application cumulative updates. That will mean we can do what we did in the on-premises era: ensure that it was going to fix more than it broke and that any disruption to the client's business is planned, minimised, and managed.

It also gives us the chance to co-ordinate with the multiple ISVs that are normal in the SaaS world and get per-tenant extensions ready, as well. Then we can test the entire application stack.

What I'd like is to be able to start a new sandbox with the new application version and my data, test it to make sure it's OK, and then promote it to production, demoting my existing production tenant to a sandbox at the same time. That gives a default option of going back on the previous production instance if anything doesn't go to plan.

I think Microsoft would be surprised how often upgrade issues happen and how much comfort it provides when that rollback option is there. Preparing for the worst is a core mentality for partners and user regarding their ERP platform provider.  

I'm not advocating for people to be a year behind; I think if the platform could run any CU from the current and previous major release that would be fine. That means that clients would have to do an application upgrade at least once a year, but once customers systems are only customised with extensions it’s a simple scripted process, right? 

2. Scale up the Support Process

The number of support cases we raise with Microsoft has gone up by a factor of thirty since SaaS came in. There is just so much that we cannot do without contacting them. That process is not always going smoothly; Microsoft has to scale up to cope.

Adding to the frustration for our support team, who have decades of experience, they usually know what will fix the client’s issue but are not allowed to do it. Even worse, they have to waste time going through Microsoft support's checklist triage process, often knowing that customer is in growing crisis. It doesn’t seem to change if it's raised as Business Critical. Don’t get me wrong, there are some good people trying hard at MS, just not enough of them and and not enough who can resolve a case without escalation.

We need the new Business Central certifications immediately. And then we need Microsoft to enforce the Business Central gold and silver criteria. The partners who have fifteen certified people, which is what it will take to get gold, should have direct access to second- or even third-line support at Microsoft. Ultimately this will create more space in first-line for those that need the 101 help. The important part is that everyone will get their issue resolved faster.

I know it is a symptom of success to have so much demand, but we want to perpetuate and multiply that success, don't we?

3. More Telemetry

The application insights introduced last October, are a great start at useful telemetry. Once you get your head around the query syntax, there is much helpful information you can drag out.

Trouble is it's nowhere close to what we've been used to getting. The last couple of years using Azure SQL we've been spoilt, being able to identify performance issues and hone in on their particular causes. That performance tuning is even more critical on SaaS, where it might impact other tenants. But even with the new insights, we're back to early SQL days in what we can see.

4. Introduce solution design reviews again

I remember Marko Perisic talking about how adopting a 'break fast, fix fast, approach' was the cloud norm. I cringed then and still do. The trouble with doing that in ERP land is the matter of fixing the transactions that have gone through while it was wrong. Even if the system is only wrong for seconds, you've introduced crap data that takes longer to correct than any code. I have a saying I've used repeatedly over the years: the fastest way is right the first time.

Especially at the application level, it's hard for Microsoft to understand, let alone test all the possible scenarios for a product as mature and sophisticated as Business Central. Automated unit tests catch lots of obvious issues, but we should stop pretending that the even twenty-thousand-plus scripted tests we have now are enough to find every problem. Specifically, there seems an issue with a lack of testing scenarios at scale. One serial process works fine, but when twenty users are all doing that process simultaneously, what happens then?

I believe that lots of issues can be traced back to the core design, I'm afraid. Not many people who have been in the partner community consulting or on the development front line for long would have signed off the design for the sales batch invoice posting process in r15, for instance.

We need to get back to the situation in which the programme managers updated key groups like MVPs, ISVs, and others on what they were intending to achieve and how they were going to approach it. Microsoft used to call these Solution Design Reviews, or SDRs. They were our opportunity to tell them what other aspects they hadn't thought of, or simply that is wasn't a good idea and why. That saved a considerable amount of coding and embarrassment.

Some imagine that the MVPs for Business Central might be playing this role already. But as an MVP I can dispel the myth. We might have more one on one conversations with some of the programme leaders, but today there is nothing formal that I'm aware of. If that started again, I don't know one of us that would not be keen to constructively contribute.

5. Introduce an early adaptor programme

Today we only get the details of what the product team are changing and how it will work in future when we discover changes that are potentially already are live at a SaaS customer. Worst case, it's when their processes have broken. Then we see all sorts of pressure on Microsoft and partners to fix it fast. That can lead to yet more unintended consequences if reactive changes are deployed without time to thoroughly think them through.

We need a SaaS-equivalent of the Windows Fast Ring testing programme where we can do what we used to do with the so-called TAP programmes back in the on-premises days. Test it with willing early adaptors who understand the risk of that early adaption. Believe me, as a veteran of ten plus TAPs over the years, fixing one customer is easier than fixing thousands and soon tens of thousands, all of which use the software differently.

6. Adopt a lean extensions strategy

I'm told that the Yodlee banking extension doesn't work with any bank outside the US or Canada, so why is the extension installed by default in the GB localisation? Why do I have the GP migration extension installed in my clients that have never had GP? 

It might be a bit old fashioned, but I'm a firm believer in not having installing anything you don't need installed. So-called lean principles should apply here, and Microsoft can do that but don't. As evidence, consider the UK's making Tax Digital extension, which is only in the GB version. If they can do that, why are customers getting frustrated by seeing stuff they cannot or do not need to use?

The installation process through AppSource is fast and simple. Microsoft and their preferred third-party add-on's should be installed this way just like everything else. The System, Base, and Localisation extensions only please, Microsoft.

I'm assuming that the mess of every extension being published into every sandbox is going to be sorted; it compromises the integrity of sandboxes if they don't precisely mirror the production tenant.

7. Finish the Admin & Automation APIs

Business Central is a volume play. And we partners don't want to sell one system, we want to sell hundreds, especially in a subscription world where we make cents per user per month. That means we need to automate as much as possible. The days of expensive people sitting for a day and configuring Business Central must be doomed. We need to be able to push a key.

In our case, we've deployed automation to our internal Business Central system to automate our own operations as a Microsoft partner, just as we would do for a client that approached us with the same issue. When we enter and release a sales order internally for a new Business Central sale, a custom extension in our system creates an Azure Org if one does not already exist, add Business Central licences plus much of our IP, and even populates data configuration packages.

But our technical solution is just a bit short. The APIs that we talk to (three separate sets) each have one or two missing criticals. That means we are so close but still so far. The one that is the most frustrating for me is not being able to install from AppSource via an API. Ironically, it means we can install custom per-tenant extensions easier than we can the pre-built ones. That mean that we're putting fewer extensions into AppSource; why would you create yourself extra work?

8. Fix the licencing

The FLF licencing that underpins any on-premises Business Central solution is not fit for purpose. I believe it's still a DOS-based system that creates the licences, and any changes are so fraught with risk that Microsoft avoid them unless forced.

For goodness sake, Microsoft, let Business Central connect up to an organisation's AzureAD and check the licences there instead of the atavistic FLF file upload. Most organisations have single sign-on setup against AzureAD anyway and, if they don't, they should. This way, all of the licences can be sold via CSP, and the dual-use licence rights would be automatic. It would also do away with partners adding their on-prem clients to Partner Source Business Centre, which surely cannot outlive the demise of Partner Source by long?

Hard to do? Lots of investment against a dying need? Well, Microsoft are the ones that keep showing the IDC predictions that the on-premises/private cloud market is going to be a flat-but-still-significant part in the future. And it's not hard to do. My company created a cloud-based licencing system for our ISV solutions from scratch in nine months with just a couple of people working on it. The payback to Microsoft for doing the same would be weeks, based only from the reduction of queries. Come on, you've been kicking this can down the road for years. Time to finally sort it, please.

9. Monetisation for AppSource

While we are talking about licencing, we need it for ISV solutions as well. That means Microsoft should collect the money for solutions installed from AppSource, as they do for their core platform and for third party solutions installed from the Azure and Office Marketplaces today.

I've already intimated that we're spend time creating our own licencing and monetisation system, but so has every other ISV. And they are all different. To say that's going to be super confusing for customers is an understatement.

10. Fill the functionality gaps in the roadmap

In the last few years, the roadmap has been very platform focused. The whole platform and developer experiences have been completely revolutionised while the functionality has been tinkered with. I understand and accept why, but that has to end soon. The competition is catching up. The strategic advantage that has come from the NAV heritage won't last forever, and not everything should be an ISV add-on.

I'll call out banking as an example. For online integration, products like QuickBooks, Sage, and Xero make Business Central look like a dialup modem in the 5G age. Business Central has not kept the last couple of years progress with the PDS2 or Open Banking standards that are lighting up FinTech here in the UK.   

I've already referenced the BCIdeas site, and I like that the top votes have been attended to consistently for the last couple for years now. That needs to continue, but it is no replacement for strategic product management. The ideas that get lots of votes are usually fairly small ‘gap fillers’ which have a value in getting done. But anything that really raises the bar on Business Central’s functional capability is usually awaiting more votes.

BCIdeas only gets votes from people using the product. How many extra sales would have been made with more of the major improvements in place? It’s not customers leaving because Business Central doesn’t do X, it’s the ones that will never start that I’m concerned about, especially as I see some major competitors raising their game in these important areas. Let's not get to the point where we are uncompetitive before starting the fight back.

The thing about a true roadmap is that it shows more the general heading improvement. It gives ISV's clear signals about where it is safe to invest. No ISV is going to fund something that might be quickly obsoleted by its platform publisher. A clear idea of the product team's aspirations for the next eight months to two years will give that visibility. Beyond that, everyone accepts too much change will occur.

Summing Up 

I hope anyone reading this from Microsoft recognises the constructive nature of this list. You won't find a more enthusiastic advocate for Business Central and the progress that's been made. There are many people I class as friends at Microsoft and I know just how long and hard they have worked to make the progress we see thus far. If you’d asked me at the start of the last decade, I’d have said what they have now achieved was impossible. But they’ve reperatedly proved me wrong. 

But I want to see another level of progress. Actually being close to perfect is more frustrating than a distance off. I hate coming second at anything because it's not winning, and with just a bit more it would have been. Business Central SaaS right now has all the potential to dominate this market like we have never been seen before, it just needs that final push to get it done.

If just some of my list gets done, then my fourth decade in this industry will be even more epic than the previous three.

Photo by David Wright on Unsplash

FREE Membership Required to View Full Content:

Become a MemberLogin
Joining gives you free, unlimited access to news, analysis, white papers, case studies, product brochures, and more, and it’s all FREE. You’ll also have the option to receive periodic email newsletters with the latest relevant articles and content updates. Learn more about us here
About James Crowter

I'm passionate about how businesses can improve their efficiency by getting process optimal more of the time. For the last twenty five years I've worked to help organisations of all sizes and types implement the ERP & CRM software that typically they decide they need when things are going wrong. I've seen that work unbelievably well and enabled those organisations to rapidly grow but I've also had some hard projects over that time where it's felt more like warfare at times.

Since 1996 (and version 1.01) I've been working with a small Danish product called Navision that's now become Microsoft's Dynamics NAV and I've also been using and consulting around Microsoft CRM since 2005. As managing Director of one of the longest established first Navision and now Microsoft Dynamics partners I've been involved in the complete history including numerous product councils and system design reviews. It's my privilege to know many of the key Microsoft executives and product designers and have insight into both where the products are now and their future direction.

More about James Crowter


hlamb's picture

Very good list, fully agree, James!

maique00's picture

Fully Agree

Thanks James.

ksebahar's picture

I would say that your #9 and #10 are the top 2 things that need to be done to provide the greatest improvements to the long-term product success.

lee.mills's picture

These two things are weaker than the competition and really seem to struggle to gain importance. I think sometimes the glossy things such as AI takes precedence over the essentials like banking and it's the product that will suffer as they are simply essential.

Bizzbrain's picture

Very good post James.
I fully agree with all items in the list.
Your appeal is already a month old. Any sign Microsoft is acting upon it?