Support @ Scale

Hey friends,

As you are all a bunch of industry insiders, I am curious to hear your take on a problem that has been the motivation of my career in the support segment of the hosting industry. We’ve all seen it, but I think few of us have ever taken the time to articulate the problem from beginning to end. This is the problem:

To start a company, many (if not most) will immediately focus on being competitive in pricing. At a small scale they are able to provide support for their customers within reasonable timeframes. However, due to their competitive pricing, their customer acquisition will consistently outpace their financial ability to scale out their support team in an equivalent fashion. What occurs far too often is that they continue the downward spiral until they feel backed into a corner, and eventually sell out. This is the point where public sentiment of their company drops significantly, see EIG brands for example.

The problem continues to exist because the price point in the market is continually brought down, forcing new companies to feel obligated (whether they are or not) to compete still, who inevitably fall into the same trap. It creates a bit of an industry loop in which companies either choose not to compete on pricing, or they continually fall in the same fashion.

There are many ways that companies attempt to solve this problem. Outsourcing to underpaid support workers is a common one. My question to you is this:

What do you think is the answer to this problem? How do you see a company pulling themselves out of this trap once in it?

On a side note, yes it is a problem that is near and dear to me with MXroute (one which I WILL solve), but this isn’t specifically about my company or any company that I work for. This is specifically about an industry trend that we’ve all witnessed time and time again. Whoever solves this problem may well be responsible for dramatically positive shifts in the industry as a whole.


I’m guessing that you’re trying to tackle the ongoing support load @ MXRoute. No harm in that — I’m actually glad that you brought it up.

I guess that the current market (one where prices continually drop) is super saturated and standing out is quite difficult. Support is definitely something that people expect to be good and is a major selling point — but like you said, dropping prices is making it hard to support everyone while making a profit.

The prices you charge, for example, are quite low. This attracts customers — yes — but the price of support doesn’t go down as you scale up. It increases and there is probably no way around it unless we end up having a smart enough AI to answer the same questions over and over again.

There are actually some chat bots that do it already, but I still prefer speaking to a real person. The interaction is what sets a company out from another; it’s the difference between a company that only has ticket support + a knowledgebase and one with live chat that charges a few more dollars every month.

What I’m trying to say that scaling is hard — customers will demand good support and will continue to do so in the future. It’s certainly not something that can be fixed in one day but generally support is proportional to the price. For example, with MXRoute, I wouldn’t expect 24/7 support for the price you charge. Customers should not expect those unreasonable times and if they want something better, they can move to someone with a SLA. (I still expect decent service though lol)

tl;dr: I’m perfectly happy paying more for better support. I’d much rather have someone local speaking to me for a few extra $ rather than an outsourced call centre representative.

Side note: I’ve never operated a company that has had such a problem, so everything I’m saying right now is based purely off my opinion.


I really like VirMach’s approach to keeping support costs low –

  • Any heavily discounted package/sale item is a completely unmanaged service (and is made blatantly aware of such upon ordering)
  • If a customer opens a support ticket for a no-support package and the issue lies on the customer’s end (i.e. not a fault with the provider’s infra), then the customer is billed for the interaction (need to appropriately disclaim this before allowing the customer to open the ticket)
  • Normally-priced packages or high-spec’d packages get normal support
  • Option to pay for premium support, if desired (maybe enforce a best effrot SLA w/o premium support add-on and a 72-hour SLA if added)
  • Provide a method for the customer to receive “community-aided support” for things that can be worked out w/o your intervention (VirMach’s happens to generally be LET, but doesn’t really matter where as long as a large user-base is present)

Just my thoughts.

On a side note - is Louis still helping you out?


This is a pretty good option. Companies that use your services will generally be willing to pay extra for good support with a guaranteed SLA. It’s similar with open-source software… Companies are often willing to pay for guaranteed support, and a lot of open-source projects end up either offering premium paid support, or corporate sponsorship.

MXRoute’s Slack seems to work well for this. I’ve helped a few people out there (basic questions like how to use cPanel’s API to create new email accounts, where to see rejected emails in cPanel, basic DNS questions, things like that). Forums are better in a way though - You can more easily search and link to previous threads.


Yeah, I joined it but I don’t really use Slack at all so haven’t kept up on it in months. I was active on their Discord while they had that in place, but I understand the reason behind the move to Slack to be more professional-oriented.


He is. Though I’m thinking well beyond MXroute and trying to take a look at the whole of the industry. I propose that this is a problem almost every growing company runs into, they just run into it at various stages. I want to be someone who can create the solutions that prevent these problems from occurring, for anyone. I think this is one of the most important problems to solve.

Support always goes to shit. Comcast is the inevitable result of growth, if taken far enough. It’s time to make it not quite so inevitable.


I’m most of the times very satisfied wirh Amazon’s chat support ( usa) despite some may pick up the phone or have English names ( I guess).
Why mention something as obvious as amazon?

I live in Latin America and I honestly don’t care where these guys are from.
They always solved every single issue I had (maybe a single case of someone being apathetic and had to click that the person did not solve my issue).

In my case what I am amazed by is that regardless of where they (support is) they are super efficient, courteous and know what they do.
How they managed to train people from abroad in case they do?
No idea


Have a friend ( usa) that was with godaddy and explained him that despite he liked the phone thing he could just open a ticket and I believe he is very happy with his new provider.
This new provider is not as big as godaddy but much more personal and set him out nicely.


In my experience the thing that sets Amazon apart is not the quality of their support but the freedom they have. I know for a fact it is (no offense intended) an Indian guy in a big call-center/support shop reading from a script sheet, but he actually has authorization to take action there and then and he will use that authorization. My best experiences with Amazon basically go like this:

Me: I didn’t receive this item.
Amazon: Ah, that sucks, can you give me the order #?
Me: Order #
Amazon: I can’t tell what caused the problem from my end but I’ve authorized a reship of the same item, it should be with you by X date.

It’s the same for returns, refunds, discounts not applying correctly, doesn’t matter what the problem is they focus on solving the customers problem immediately and not finding the cause or placing blame in the moment. I suspect they absorb many losses through this strategy, because they refund, replace, etc first and ask questions later.

This is equivalent to every level 1 ticket being put straight through to a level 3 tech/sysadmin - the problem is going to be resolved, the response is going to be as clear and direct as possible, your experience is going to be miles ahead of the level 1 response on average.


Lovely said!
Amazon makes me want buy from them again.
I had ( blame me and my bank) issues with the first attempt of purchasing tje marshal headphones.
Later on when I fixed the issue I saw they had raised their price.
I was a bit silly and told hem if they could preserve the original price and they… ( this true) transfered me to the price match department.
Price match department… I have no clue if I was always speaking with the same person but wow amazon what a nice way to make me buy from you again.
Well they could not price match it.
Found and alternative ( Brown instead of of black).
But… They had no stock ( as despite what one of the many hits told me which was to pay the quickest shipping that they’ll refund me it would not make it on time ( my brother would be back to Argentina). Obviously felt great that they would refund me their most expensive shipping option but the bad news was that it would male no sense as he will be back ( my brother and I won’t have my God damn headphones).

So what I’ve done?.
Camel camel had some issues and I could not use them to have a good perspective so the little obsessive in me kept kept hunting what I’ve wanted till I finally found what I wanted with free shipping.

Yes they created a dependency of sowmone who is 10 thousand kilometers from where they sell their stufzz but they’ll open a new warehouse in Brazil :wink:

At scale what I think they know is how to create dependency haha

1 Like

I appreciate all of your thoughts on this. These are my most recent thoughts on the topic:

As you can tell, regularly experimenting on this is kind of my thing right now. You have to be willing to fail, and be willing to bail out on a bad idea when it doesn’t solve the problem.


Raise your prices for new customers until you are financially able to get quality support, once you reach equilibrium, reduce prices and so on.

Also, crowd support. If you’re tied up with presale or general questions not specific to a client, crowd support can be a big help.


Instead of doing multiple things, with mediocre value, in the same daily timeframe, one should focus in excelling at one thing, and one thing only, and stay in that earned top spot.

However, we are curious humans, we change, and we get bored of routine (no matter how much we worked to obtain that routine) - maybe a fault in our organic design.

1 Like

Since I try to refrain from generalizations I have no particular advice.


Most Valuable Comment of the Year.
Not disappointed.


Why the rush? Year 2019 is far from over. Rewards should be given next year.

1 Like

Thank you, sir.

can’t agree more. it’s not about fixing the support strategy, it’s about adjusting the business model.

start being more selective about customers, quality over quantity and raise prices to match the good service you deliver. try to get a feeling for the relation of scaling speed vs pricing development.

convert your customers to promoters (I think you can say that you did that already for mxroute) and if your support gets better raise prices again, because happy customers will understand and stay anyway.

you need to turn your good reputation which you worked for really hard into revenue finally. and that’s even in the interest of your customers, as they surely want you to continue to provide exactly that.

ramnode comes to my mind, which had cheap pricing and a lot of coupons and stuff in the beginning. they also earned themselves a very good reputation. and they simply stopped doing that crazy discount stuff and also just did not increase the offered hardware specs on the servers, while a lot of competitors did exactly that (offering better specs for same or less money).
still ramnodes customers stayed and praised them…

tl;dr; jump off the race-to-the-bottom train. done.


It`s quite interesting to follow along this thread :slight_smile:
There are many good suggestions out there and I can’t really think of anything to add right now. But if I can, I’ll leave it here.

There once was a provider (can’t recall the name) that offered support in a way that may be a bit similar to Virmach:

On one hand they offered super budget, price competitive, unmanaged plans with no technical support. Only Sales and a Knowledgebase for technical problems. Optionally you could create a technical support ticket which would cost 3-5$ (iirc) where your inital question would be answered and you could ask for clarification of the support agents solution until satisfied. Once it was solved or no reply/question by the client for 48 hours the invoice for the ticket was generated. If after that the client would re-open the ticket it would cost like 1$ per new reply by the support agent. Any new ticket again did cost like 3-5$. Maybe it sounds like a rip-off, but for an unmanaged solution I found it to be quite the interesting solution.

On the other hand they offered “normally” priced plans that included normal support + offered sorta “hand-holding” for an individual monthly fee upon request.


As someone who is looking to launch a venture into a crowded market place I hear you.

This market place is crowded but there is a USP to the product I want to sell. I don’t think without a decent marketing budget or the ability to put money into the business in the early days there is even any point in launching. I don’t mean that in a defeatist way, I mean in a pragmatic way.

My gut feeling is the Amazon way could be the best way to do it, but you need to have the finances in place to support you to do it.