The Dysfunctional Dumpster-Fire of Information Technology in VC-Backed Tech Startups

IT is a dumpster-fire in many organizations. Don’t take my word for it—ask your employees. The IT dumpster-fire that’s near and dear to my heart is in VC-Backed Tech Startups. Some of this may also apply to other startups or established companies, or even government agencies. But this isn’t about them—this is about VC-Backed Tech Startups and their journey to fully dysfunctional IT. Technology can be a massive force multiplier when done well, or at least adequately. But done poorly, technology is a force divider that drags down every person in your organization.

1 – In the beginning, Founders.

Tech founders start with an idea, to build a product or service. IT isn’t something on their radar. They almost always start with personal laptops. They start building a product and working on their pitch for funding. At some point they’ll decide on a name for the company, and one of the founders will register a domain. Now they have a domain, so they need to want to start using it for email. One of the founders signs up for a small business plan at Google Workspace or Microsoft 365. They need some cloud hosting, so one of the founders sets up an AWS, Azure, or GCP account on their credit card. Eventually, through bootstrap or funding, they start to hire some people. One of the founders orders them a laptop. Whichever founder setup the email service creates their email account. The cloud-hosting founder gives them access to cloud resources, usually as an administrator because they don’t know what they’ll actually need and don’t have time to deal with it. New hires get their laptop on their first day, hopefully, and then start trying to figure out what software they need to install, and which founder can buy it for them or give them access to it or whether and how they should expense.

2 – Engineers Build Things.

Some of the first hires in a tech startup are usually software engineers. These engineers usually complement the founders’ skills and start building parts of the product that were outside the founders’ wheelhouse. Inevitably one or more of these engineers will have something resembling IT experience in their history, so they’re now the unofficial IT Architect. They’ll set up remote access or VPN to the cloud hosting. If there’s a physical office, they’ll set up a network. They may take over some of the cloud and email administration duties from various founders, or just be the single point of knowing which founder admins what. But their primary responsibility is still building product. They don’t spend much time doing tech-support for other employees because everyone is technical, and those people can Google the problem themselves.

3 – Money Changes Everything.

At some point, the company will have funds to hire more people. Maybe they got a good seed round, or maybe they are already generating revenue. Either way, they invest it into growth by hiring more people, and some of them aren’t technical (looking at you, Sales, and you, G&A). Unofficial IT Architect isn’t happy about babysitting new Sales rep while he installs and configures Outlook, and this IT crap is starting to affect sprint velocity. It’s time to hire an IT person.

4 – A Guy That Can Do It Cheaper.

Unofficial IT Architect is sick of doing low-level IT tasks, so somebody hires a low-level IT person to take over that work. The title varies, things like IT Specialist, Desktop Support, or maybe even (junior) Systems Administrator, but the work is pretty much the same: create mailboxes, set up laptops, help users reset their passwords. New IT Person reports to whoever has time to deal with them—could be a founder, could be somewhere in Engineering, could be Finance or HR. The company keeps growing, through revenue and/or funding, and New IT Person gets a lot busier. New IT Person says they need more help. Whoever got stuck with managing New IT Person starts getting complaints about things not getting done, and Whoever doesn’t have time to deal with complaints about emails and laptops. At some point, the founders decide they need an IT Manager. They may promote IT Person to IT Manager and hire another IT Person, or they may hire an IT Manager. Either way, they go cheap because emails and passwords aren’t that important and what they really need is more engineers and salespeople.

5 – The Rise of Shadow IT

The company has been hiring a lot of people--mostly engineering and sales, but also product technical support, some people that do accounting stuff, a marketing person, and even an HR person! And the small IT team may have added another IT-Desktop-Support-System-Specialist-Administrator too. IT is pretty busy with laptops, emails, and passwords, and they don’t want anything else on their plate. And IT isn’t doing that great with laptops, emails, and passwords, so nobody else wants anything more on IT’s plate either. Sales has outgrown their customer spreadsheet, but they can afford to buy a CRM subscription, so whoever in Sales has made their own reports in Salesforce before goes and signs up for a Sales Cloud subscription. Marketing wants to send out some email blasts, so the marketing person signs up for marketing automation platform subscription for himself. Accounting already set up their own software a while ago. The founders decide it’s time for a real website, so the marketing person hires an agency to build it and forwards the agency’s email about some “DNS” thing to IT and Engineering so one of them can figure out what to do. The Tech Support Manager has signed up for a ticketing system, and Engineering has their own development stack, cloud hosting, and VPN.

6 – Revenue Is King

Every part of the company has their own business apps, and none of them are integrated with the others. The biggest pains are in Sales and Marketing. Managing Salesforce is beyond the skillset of Sales Person who signed up for it, and Marketing needs CRM data integrated with their Marketing Automation Platform. Tech Support pipes up and wants CRM data too, and eventually someone figures out that CRM needs licensing data from Engineering for them to sell expansions and renewals. Accounting may want some integration too, or may stay with spreadsheets for a long time because they don’t trust the numbers from CRM and Engineering. All of them come to the same conclusion: this mess is affecting revenue. And like most startup revenue problems, the fix is to throw more money at it—in this case, by hiring someone. Other technology isn’t affecting revenue, so this is scoped to just CRM and its close, revenue-impacting friends. Whether they hire a manager first or technical people first will vary, but eventually you end up with a team called something like Business Applications or Business Systems ensuring that revenue-generating systems actually work. And they report to the Sales branch of the org chart, of course, because this technology is a revenue problem, not a technology problem.

7 – Choose Your Own Adventure

There are several ways that IT can go from bad to worse from here, but they all boil down to having a mismatch between the leaders, their goals, and the organization structure and a lack of vision on the importance of technology to the success of the business. Some examples follow. I’m sure there are more, but these are the ones I’ve seen. Which dysfunctional path an organization chooses depends on the current priorities of the company, but, SPOILER ALERT, none of those priorities have anything to do with email, laptops, or passwords. Also, you can pick more than one or mix-and-match pieces.

7(A) – Seven Key Metrics for No Tomorrow

There are a few ways that a company gets here. It could be cost-driven—is IT being efficient with the resources they have? Are there underperformers that need to be put on a plan/replaced? Or it could be performance-driven. What is IT spending all their time on? How much time are they actually working? Are there underperformers that need to be put on a plan/replaced? Of course there’s some overlap between cost-driven and performance-driven, because they’re really the same thing when it comes to how the company views IT as a “cost”. We’re adulting hard now and need to make data-driven decisions. To get metrics, the first thing to do is make sure that every request comes through a ticket. No direct messages or emails, no hallway conversations or walkups, EVERYTHING goes through a ticket. But ticket counts aren’t the only data that budget-makers want to see—they want to see what the problem areas are, and what teams are burning more support costs. So all of that needs to be fields that the user submits on the ticket. A good ticket should have a field for the problem area—there’s a limited set of problems/tasks that IT deals with, right? And the ticket should have a field for the Application. And there should be a list of teams—accounting should have a list to copy/paste into the ticket form, and those don’t change very often, right? Once you get all of the fields you need in the ticket form for your metrics, it shouldn’t take a user more than 5 minutes or so each time to fill it out, especially once they learn the list of applications. Was Microsoft Word listed as Word, Microsoft Word, Microsoft Office, Office 365, or Microsoft 365? They’ll figure it out—what matters here is accurate metrics. So now we have some data, so let’s look at some important metrics. Let’s start with Mean Time to Acknowledge/Respond. Sometimes it takes a few days for IT to respond to a ticket. Wait, we can fix that with automation! Auto-response on ticket creation gives us MTTA under 1 minute. Done! Now our IT is now awesome. On to the next metric, Mean Time to Resolution. MTTR is big, but IT says they are waiting for users to respond, so it’s really the users’ fault that MTTR is high, and IT can’t do much about it. Ignore MTTR. Let’s look at ticket closure metrics next. Sam closed a weekly average of 37 tickets in a row, but Pat only closed a weekly average of 9 tickets. Sam is new and mostly does password resets, and Pat has been spending a lot of time troubleshooting problems that Google didn’t have an answer for. The data says we need to put Pat on a PIP to improve that closure rate. We also need some metrics on user feedback, but we’ve only gotten 1 complete response to the 13-question ticket closure NPS survey year-to-date and we don’t feel that it’s enough data to be actionable. Add more metrics as needed to take up space in the PowerPoint presentation where these eventually land.

7(B) – Leadership Through Project Management

So you have a Business Apps/Systems team that does really important computer things like CRM that help bring a lot of value to the company in revenue and growth, and you have an IT team that does boring computer stuff that doesn’t bring in any revenue and just costs more and more money. The Business Stuff team does big projects and has nice Gantt charts for them. IT just replies to tickets, eventually, sometimes. The Business team complains about IT being a speedbump whenever they need something from IT, like email or authentication. The solution? Put IT under a Business Apps/Systems leader and call them the VP of Information/Business Technology. Divert some of the email, laptops, and passwords work from reactive/ticket-based tasks to managed projects. Dedicate some portion of the IT staff’s hours to working on managed projects instead of tickets so that projects can maintain velocity. The Business Stuff team is happy, Sales leadership is happy, the CEO is happy, everybody is happy, except for the-people-that-have-to-deal-with-IT, aka The-Entire-Company. The original IT workload hasn’t gone away, and the resources processing the workload haven’t increased. The workload has just been diced up differently, and now most IT issues take even longer to resolve because projects are preempting ticket work.

7(C) – Make Security a Priority

There are two ways a company usually gets to this point. Either 1) they start losing deals because they are missing security checkboxes like SOC2 and ISO27001, or 2) they have a material security breach. Since both reasons are generally about the product, they start by hiring a Product Security person. Product Security dives into security issues with product, and either already knows that a compliance project is way out of their scope, or they figure it out really quick. IT is useless except for laptops, email, and passwords, so the leadership figures out they need a CISO, with experience running a big security team. They may bring in a consultant/vCISO to help get the compliance checkboxes in place, or they may decide that this is actually important and pony up a management headcount for the CISO. As it turns out, there’s a lot of stuff about laptops, emails, and passwords in this security compliance project, and the new CISO seems to care about those, unlike whoever currently manages the IT Manager. And before you know it, the IT Manager reports to the CISO to “make security a priority”, totally not because the CISO had spare cycles and some real C-level was sick of dealing with IT. Anyway, getting these security compliance checkboxes is important, and the CISO’s annual bonus depends on it. The CISO’s bonus doesn’t depend on Corporate Bro getting his password reset so he can update his Salesforce opportunities or the new engineer getting a blingy new-hotness laptop. That stuff can wait, security is the priority now. IT was dropping stuff left and right anyway, so the company is used to that. For now, all new spend goes to supporting the top priority/bonus KPIs, security. The CISO builds a security team, they get some compliance checkboxes done, Sales leadership is happy, the CEO is happy, everybody is happy. Except for the-people-that-have-to-deal-with-IT, aka The-Entire-Company. Business Apps and Engineering avoid dealing with IT, and for everyone else, IT will get to your ticket eventually. None of this is to say that security isn’t important, but that’s another dumpster-fire for another article. TL;DR: putting operational IT under a security person doesn’t make operational IT work better.

7(D) – A Guy That Can Do It Cheaper 2: Mediocre Service Providers

So your on-staff IT isn’t that great, they’re really expensive compared to the cost of a single laptop, and somebody “knows a guy that can do it cheaper”. The Managed Service Provider offers Service Level Agreements too, which internal IT has never been willing to discuss much less commit to. They may even have options for 24/7/365 support. Maybe you just outsource part of IT like laptops and on-boarding and keep some people on-staff to deal with fixing printers and supporting some internal applications, or maybe you yeet the whole IT staff and outsource it all to an MSP. Managed Service Providers, like most consultants and professional services, provide the same service to a wide variety of organizations. They standardize their offerings so that their staff only has to know how to perform an operation, such as creating an email user, a set number of ways, so that their staff can do that operation for every client instead of having to know a different process for every client. Abstracting that a bit, the MSPs sell square pegs. They typically can scale up or down the square peg to fit the un-square hole that is your company’s IT needs. Some businesses don’t have any idea what shape their IT-hole is and are happy to adjust how their business operates to fit the MSPs square pegs. Technology startups tend to have specific IT needs, whether or not they are actually met, and the shape of their IT-hole looks like it’s an area defined by the intersections of multiple higher-order differential equations. You can jam a square peg in that, but there will be massive gaps that aren’t covered, particularly custom integrations and workflows. But that’s OK, because most of the company is already accustomed to working around IT’s limitations, and now you get a live person who can’t solve their problem to call them back within 24 hours, and they get $1 off their next pizza if they don’t get that callback within the SLA period. And that 24/7/365 support, that’s just to talk to the Level-1 person that can’t solve your problem—the people that handle Specific SaaS Product troubleshooting work 9-to-5 Eastern, weekdays. Well, unless your problem was something simple, but if it was that simple, could it have waited for normal business hours? Magic 8 Ball says “signs point to yes”. Are your employees happy with the 24/7/365 Square Peg? Magic 8 Ball says “Outlook not so good”. While you’re at it, outsource your cheap/disposable Brother multifunction printers to in favor of a leased big printer/copier so you won’t have to dedicate all those IT resources to ordering and changing toner cartridges once or twice a year.

7(E) – This Is Why We Can’t Have Nice Things

Maybe your runway is getting a little short, or maybe your investors are laser-focused on your EBITDA margin affecting your valuation. Either way, that means cost-cutting across the board. Of course, you’ll have to lay some people off, but you’ll also want to look at everything else that you can do cheaper. IT makes a lot of purchases, so that’s a great place to look for cuts. Stop replacing laptops unless they are broken beyond repair. Buy cheaper laptops when you need new ones. Consolidate similar software, particularly collaboration tools, with the cheapest common denominator. You can save thousands of dollars a year and only lose an hour or two of productivity a week from most of your employees. And all those consolidations may only take hundreds or thousands of person-hours to complete. That’s fine, because we’re focused on cutting costs here, not building product, expanding markets, and growing revenue. For more cost cutting, see 7(D) about outsourcing to a Mediocre Service Provider for ideas about eliminating most of your IT staff. Note that “moving to the cloud” isn’t in this section. Cloud is awesome, especially for development and scaling, but it’s rarely a cost saver—but that’s a different dumpster-fire for a different article. TL;DR: giving your remaining employees crappier technology will make them unhappy and less productive.

8 – Purgatory

You have arrived at your destination. Your IT is completely dysfunctional, and your employees hate it. On your next regularly scheduled employee engagement survey that you haven’t done in a while, a lot of employees write in something about IT in the “What can we do better?” comment box, some even with colorful language. Now starts the endless cycle of denying problems exist, shifting blame, making excuses, and small changes that don’t address the root cause. “Our metrics don’t show this”. “Users are being unreasonable!”. “Pat is on a PIP now; we can replace him soon”. “We need budget implement X tool and Y people next year”. “The MSP is meeting all their SLAs”. “It’s not that bad, and we’ve cut IT costs by 17% this year”. Many startups stay like this until they fail or get acquired, some are even very successful despite their dysfunctional IT—after all, they can work around IT.

9 – Stop the Dumpster-Fire – I Want to Get Off

There is a better way. IT doesn’t have to be a headache, a speedbump, a roadblock, or a complete dumpster-fire. It can be fixed at any stage in the stories above. The first step is an acceptance, acknowledgment, and daily self-affirmation in front of a mirror from the CEO, Founders, or whoever is currently running the business—repeat after me: Technology is important to the success of my employees, and the success of my employees is important to the success of my business. Once you have accepted that technology is important to your success, you need to invest in it. This is going to cost money, but the payback is all over your balance sheet, from revenue to payroll. Start your investment by consolidating business technology under a single leader that shares the vision of employee and business success. Put laptops, emails, passwords, CRM, ERP, HRIS, HVAC, and every other technology other than product under one team that can make them work together instead of working around each other. The title for this leader will vary, particularly as an organization grows, but this leader’s peers will be your other business leaders from Sales, Engineering, G&A, etc. You can even throw security under them and call them a CISO if it makes your investors happy, but remember the stuff about technology and success—don’t just hire a security person. This leader should never say “that’s not our responsibility”, and proactively offers “how can IT help?”. This leader will treat new technology requirements as a challenge and opportunity instead of saying “we don’t have anyone that can support that”. This leader will strive to understand how the business works, how employees do their job, and seek out ways that technology can make both more successful. They will be driven by customer satisfaction, and actively seek and destroy processes that impair employee happiness. They will constantly look for ways to integrate systems because copying/pasting between systems is time-consuming and error-prone, and they’ll look for opportunities to do broad integrations that span multiple systems. This leader will cross-train employees and tear down silos so that troubleshooting multi-system problems isn’t a game of hot-potato between teams. This leader will have done some or maybe even all of this at a startup before. Even if they haven’t done it all, if they have a solid technical and management background, share the vision of employee success, and embrace the rapid change of a startup, they will grow to do it all. This leader will have enough technical skill to call BS when bad data is being presented for a data-driven decision, and will build relationships to gauge success up, down, and across the organization. Will this leader “reign in” your IT costs? Probably not, unless you were seriously overpaying for something. But technology is cheap. People are expensive, and turnover of people is expensive. Technology should be working to make the business and its people successful.

About the Author

Jon Lusky is a serial tech startup veteran that has built and led highly successful technology teams. He’s helped the teams at multiple startups use technology as a force multiplier to springboard their growth from initial funding through M&A and successful exits. Jon is currently available to do this for your business. https://www.dhirubhai.net/in/jlusky/

?

?

?

?