First let’s dispel a myth that automation is just about software, systems and robots doing the work. This is how people traditionally think of it. Automation is also about people operating in a commoditised or procedural way. This article is going to address various types of automation.
People are very ritualistic, process-driven, free-range and organic – they’re real. They don’t have the constraints of a system or software which are very rigid and structured, yet we push ourselves to work constantly often like machines. We have designed ourselves into this situation thanks to our western work ethic.
The problem with this it is easy to become a human automaton, doing without thinking. Your perception taught to you by schools and your parents is that conditions shape your success and your chances are better if you work hard. This is wrong, it’s good decisions associated with hardwork that shape success. Hard work is the accelerant.
The paradigm we live in frames this work ethic. Firstly, we’re in a very high debt and tax-paying society. We think we need to grind to deal with those two major ‘recurring’ drawdowns on our wealth.
What we need to do is wakeup to the liberty of automation. Freeing up our time for new goals and ventures.
Types of automation
There are three key automation methods:
People should be considered part of this picture. At the commoditised end of work tasks are very automatic, structured. Think iphone assembly line worker. This is automation from the business owners perspective. Results are very predictable from time and quality perspectives. It certainly isn’t automation for the line workers perspective. Outsourcing is an aspect of this. These service providers combine systems and people to automate tasks.
Knowledge worker <-------------------------> Commoditised worker
Head of Strategy <-------------------------> iPhone assembly line worker
Machines and robots is the next stage of this automation. For example Foxconn use robots to replace line workers when building iPhones.
Software can be seen as the knowledge worker version of Robots. It can replace human tasks like accounting, SEO, payments processing, marketing, copyrighting and soon it will doing your tax returns. Artificial intelligence is providing more solutions to roles on the left of the spectrum diagram above.
Automation is about rituals
Rituals are very important. Automation is a long term project. You chip away at it. It’s not something you can do overnight and is definitely an ant versus grasshopper story. It takes patience and perseverance. If you can put up with the front-loaded pain and invest time in automation, you will get the rewards over time.
Automation suits people who can delay gratification, but not everyone has a personality compatible with that. It does take a bit of effort to work despite a natural inclination to be satiated immediately. I think that’s why a lot of people just do stuff manually without really taking the time to try and automate. They never get around to it
Pick one idea in each area to action at a time
Experience Refinement : Customer Advocacy : Marketing Flywheel
What we do to try and deal with this problem at Saasu is look at the different areas in our own business. We have a list of the ideas we want to do within our strategy and we try to pick them off one at a time. The problem is that we break that rule a lot. We’ve got to pull ourselves back and realise when we’ve taken on too much. It’s natural for humans to take on more and do more, but before you know it, you’ve loaded yourself up. You never get a chance to actually do the work you should be automating because you’ve built yourself a mountain of manual processes. You have to say no, do less and deliver. You also need to build systems to automatically deliver for you.
I’m going to start with the obvious one, email processing. The reason I want to start with this is because I think it’s one of the big pain points for many small business owners. It may be something that’s been talked about a lot in the market, but I wanted to share with you what I’ve been doing personally, just to give you a bit of a feel for how I operate. Hopefully there’s something that you can take away. It may not add anything to what you already know, but at least it might spark ideas.
Email reprioritisation techniques
Email can really own you. It’s hard to answer everyone. It owns you because you feel compelled to answer everyone. You feel guilt when you don’t answer, and you feel vulnerable when you haven’t checked in to see if you’ve got any emails that might be important. It utterly enslaves people.
One of the things I’m doing and one of the obvious things to start with is email filters. I filtered emails into folders in Gmail for review as an example. It enables me to just click and look at those reports when I want to, rather than constantly filling my inbox with all the reports. I do this a lot across the many reports that I have coming in.
What this really does is facilitate planned email maintenance rather than interruptive email. I think most people operate in interruptive mode with email. I like to think of email more as a feed rather than a task list, because as soon as you start seeing email as a list of tasks, that’s when it really starts to own you. You inbox will stop you prioritising time basis your goals.
Email to case
Email to case is another one we use more now. We’ve used Salesforce and LaDesk for our case management and sales at different times. It doesn’t really matter what system though, you can usually auto-forward emails into a case management system. This is an email account that I manage which forwards straight into Salesforce so I can treat those inbound emails as case tickets. I do this because I don’t do service as much anymore, but there’s some quality, technical and CEO feedback things that I like to answer, to keep my finger on the pulse. So it’s really important I get that in there to make sure I do answer properly. All the good CRM systems have this type of capability.
Systems vs email for work
The next thing I try and do is shift stuff from emails into the areas where they should really be, where it’s more automatable and manageable in a structured way. Getting things out of email and into cases, project systems, CRMs, and transaction systems.
A lot of small business owners start up and do quotes from their email. That’s crazy in my opinion. As soon as you do that, you load up your email system into a purpose which it’s not really designed for. It’s just inefficient. Even short conversations like chat, quick meetings, and small questions shouldn’t be in email. They should be in a chat system like Slack or Skype. It’s best to try and de-load email as much as possible.
Scripted or canned responses
This is one I’ve got for VCs. We get lots of requests from VCs for meetings and investment interests from them. It’s annoying having to answer them all the time. I don’t want to be dismissive. So I just built a canned response for it. I might add my own little bit of context to it or change it slightly, but it just speeds things up because I can just send it straight away. I’ve got them labelling automatically in Gmail so I can go back and check them. It’s just an easy way to manage a common request that’s very similar across multiple people. I don’t have that many canned responses, but obviously for things like the CVs I receive, inquiries for investment, and things at CEO level, I use them. In our service team, obviously canned responses in systems like Gmail are used heavily.
Marketing automation is the next area. I think this can be a juicy one for small business. We’ve use systems called Mixpanel, Marketo and Salesforce. I’m going to talk about what we’ve used because I want to give you the behind the scenes view of what we’re doing.
Using big data in small businesses automatically
To start off with, you can’t have this conversation without looking at what big data is and what that means for small business. Pretty much any business now needs to be scoring its customers. People might say that that’s just about software, but I don’t agree with that. I think it goes right across industries. Even industries which are very physical need to be doing this. How do you properly allocate human resource and financial resource to new opportunities properly without scoring those opportunities? You have to and make sure your attention is going on the right things.
I think if you don’t know what big data is, this cartoon sums it up pretty well. It’s a really nasty and cold version of how you would see big data, but essentially it’s the reality. Systems are capable now of working out how likely a customer is to buy, or whether the customer is really interested. That type of data really saves a business like us from harassing a person if they’re not really interested, or alternatively, giving them the information they need if they are interested. It is respectful of customer when used ethically. We we don’t do is track lots of personal information. We are very against that. We just try to track level of interest and use of our products. We also track some basics that help our content remain relevant like our customers industry.
Using tags to segment data in systems like Saasu
Tags and segmenting data in Saasu is something we do. We run our own product, Saasu, and we use our own systems. I think we’re one of the few software companies that do that. The example in the image is a contact record of a partner. You can see the tags we run on that person inside Saasu. Beta means they get access to the beta releases of our product to test before it’s released to the market, Published means we published them on the website etc. All of those tags are really just segmenting. As we’ve put tags on the different contacts in our system we’re building up a very segmented database of contacts. When we pass that information over to other systems to create really defined groups of people. These tags are essentially metadata. It becomes quite an easy task for someone in Saasu to just export a CSV file for this data and import it into an email system, or use a connector to have that information flow across automatically.
We use a custom built notifications system to send notifications automatically or link customers to online content. Different notifications are automatically getting triggered based on online actions of customer. We try and do this in a way that’s not harassing though, so they don’t always go out. It takes certain rules to be met before these go out to ensure that people aren’t getting them willy nilly. Not everyone gets the same messages, so it’s a customised approach. You can use systems that are off the shelf like.
Segmenting data in email
Segmenting data in email systems is common place now. We’ve used Campaign Monitor above (expensive) but we are about to move to Mailchimp (easier and cheaper). Both have good segmentation. In the old world, people used to run lots of lists, which is just a really bad way to run things in my opinion. I think it’s hard to maintain lots of different lists. These email systems actually tend to penalize you for doing that now because you have to pay a lot more if you’re running lots of lists.
For us, a field in Campaign Monitor is like a tag inside Saasu. We could have a tag for each one of our variables. Then, when we want to send a campaign out, which is more of a big scattergun approach to emailing, we have a more targeted way of doing it that’s also simpler to manage.
If you really want to do automation of marketing, you’ve got to track the customer journey. This can be done in systems like Mixpanel. You don’t actually have to do this yourselves, just had to add a code to our application and websites to make it work. When people come to your webiste initially, they’re anonymous. You don’t know who they are. If they sign up to your product or service and they traffic your website for a while, you might then find out that they’re a known contact.
From there you might score them. Scoring is something you can do to work out whether you want to engage the customer. For example, as a customer starts surfing through a website and application and uses it a lot, you’ll find the score tends to go up. It gets to a point where it hits a threshold, which might trigger an event. That event might be an email offering them a deal, or it might be an email that goes off to a sales team in some business models. The sales team is then responsible for getting in touch with that contact, helping them out or converting them, depending on the business model you have. The conversion process inside your sales process is really the last step.
The important thing here is that different people need different things, so based on the scoring and clicks, they might get different outcomes. A partner of Saasu might get a brochure or an email about the partner program, while a business owner won’t get that because we quickly work out they’re not a business owner because they didn’t tick advisor when they signed up. We know they’re not an accountant, they’re definitely a business owner. There’s a lot of little ways you can make sure you’re getting the right information to the right people.
Scoring customers also enables you to prioritise. The image above is from Marketo. In the image, you can see it’s just scoring people according to activity. You can quickly get a list of people that you need to market to, have a sales team follow up on, or add other processes to etc.
Automatically Check-in with customers
The next step is checking in with customers.
From time to time I have setup an automatic checkin email we send new or existing customers. This is so I can provide some personal service or review feedback from customers. This type of Trigger Events is excellent for this because they cause something to happen at the right time in the customer journey. I like finding out what people think about the product and what they’re after in the product. This is the kind of information you really need to ask for, but I certainly don’t want to send out thousands of emails to people signing up to the product all the same. That’s just a waste of time. So I automate this email.
I wanted to touch base and personally welcome you to Saasu – thanks for giving us a go!
I’d like to invite you to contact me personally if you have any feedback while you’re using Saasu via firstname.lastname@example.org. I’d love to hear your perspective. I read and answer all the emails I get as I can get to them. For service questions our Client Managers are here to assist also via email@example.com also.
People sometimes think it’s insincere to send an automated email, but I do actually answer every one of these that comes back. I am writing an answer to every single person that responds to this email. I’d say if you’re going to do automation, you’ve got to make sure there’s an integrity and sincerity behind it. If you are asking them a question be prepared to do something about it when they reply. A check-in is a good chance for a business to find out what’s wrong with the business. It also saves you collecting information in an automated way. Not all customers reach out with problems, so this is just a way to collect as much feedback as possible.
Transition time – using automation to reduce it.
Disclaimer: I’m going to use our own Saasu app as the example a few times below. It’s what we use for financials and transactions so it’s the best way to explain by example.
Transition time is the non-productive time you spend between actual productive parts of a task or goal. For example; ducking down to the hardware store while you are building something for some bits. The travel time is transition time. It’s a waste if it can be delt with in another way such as delivery or bundling up purchases. Another example in technology is having to navigate through a bunch of screens to get to what you want in a website that you use often. It’s all a waste of time if there is a direct path available. That’s what bookmarks solved in web browsers.
Transition time is really one of my pet hates when it comes to time wasting. I see this as a big opportunity in many small businesses. Think about all of the stuff that you have to do and the time you waste moving between systems or places or things.
Too much stuff, too many systems
Take a look at this desk above. It’s just got too much stuff! I guarantee it’s probably a complete mess inside the computer too. It’s just really hard to take being efficient seriously when you’re in this situation. I really believe in minimalism and keeping things clean and lean if you want to be efficient. That’s actually the first step before you can really automate.
If you want to run an inventory-based business, for example, and you’re transitioning between inventory systems and the accounting system as you’re processing transactions, you’re wasting a lot of time.
Another example is if you’re trying to build a quote inside Word or a Pages document, then you try to have that inside an accounting system, then you try to send a quote via attaching it to email. That’s a whole lot of wasted effort there. What you should have done is just generate that quote inside the accounting system and make use of its power. That way, you’re avoiding all that transition back and forth between systems.
Another classic example is people checking contact records between different systems to make sure they match. It’s the double data silo problem, as they call it. There’s lots of examples of this, but too much stuff and too many systems is what leads to that problem.
Technology that automates transition time
We’ve got a search feature in Saasu which attempts to address the transition problem, and it does it really well. Time is lost in navigating accounting systems. Let’s say you want to see a particular customers Invoice. You’ve got to drill into an accounts receivable area, then you’ve got to search for the contact or search for the transaction. Maybe use some filters. Even when you find it you’ve still got to click down once more into the transaction.
We built search based navigation into Saasu accounting to solve this.
You just type something associated with the customer. Invoice number, name, email, stock code etc. and instantly, within a second or two it displays likely results (similar to instant search on Google). It just removes that tedious experience of finding stuff.
Trying to find things is an experience people do thousands of times a year and waste thousands of minutes on it. They do this across multiple systems. It’s even worse when you’re outside systems, say you’re dealing with paperwork. Transition time is a real killer inside of business efficiency, because it sucks such micro bits of time that it doesn’t get felt and doesn’t get dealt with. You’ve got to make sure systems have good navigation and speed of use to keep things efficient.
Transaction processing is another one of my favourite areas because there’s just huge implications here for efficiency. The first one I’ll start with is what I think is the mecca of automation, and that’s straight through processing.
Straight Through Processing
If you’re not familiar with straight through processing, it’s mainly something that’s come out of the banking industry where you have things created automatically by users. When you do a transaction or a trade in a bank, it ends up in a system. Another system generates a matching ticket, and it all kind of just flows through from that first piece of data entry.
You can have this in small businesses, but you have to invest some time to get it. It basically means there’s minimal business owner involvement because the customer is entering the transaction by going to a point of sale terminal or entering an e-commerce site. The checkout process is obviously all automated there, and it goes through the merchant, which goes through to the bank. You can get your bank data feed in to collect that data, and then you get your KPIs from your accounting system. You really haven’t had to be involved.
The main thing that’s really left is just to have a really good bookkeeper, and have them making sure everything’s reconciled and working okay. They become almost like your system controller. They’re high value knowledge workers in this situation. When they do that last bit well and make sure the system is running nicely, the rest of it will run like clockwork.
With straight through processing, there’s a lot of pros and cons. There’s different ways you can do things. People see bank data feeds as an example of straight through processing, but it’s not. What I’ve done here is put together a bit of a chart to give you an idea of where the benefits are. The top one there, straight through processing, is the best. It’s one star because it takes a fair bit of time to get set up and you’ve got to spend some money to do it. But once it’s going, it has great ongoing time value. The data delay is virtually nil, it’s all instant. The data accuracy is also very high.
|Setup Time/$||Ongoing Time/$||Data Delay||Data Accuracy|
|Straight through processing||X||X X X||X X X||X X X|
|Merchant data feeds||X||X X X||X X||X X X|
|Bank data feeds||X X X||X X||X X||X X X|
|Import CSV files||X X X||X||X||X X X|
|Manual data capture||X X X||X||X||X|
Merchant data feeds are equally just as good. There’s just a bit more of a delay because it takes an extra day.
Bank feeds aren’t actually all they’re cracked up to be because they have a day or two delay to get your bank data feeds. They’re very easy to set up, but the delay and the ongoing time cost isn’t quite as good. There’s a little bit of work there to make it work. It’s not as good as straight through processing, but it’s still excellent.
As you get to importing CSV files or manual data capture it gets less efficient. Obviously, there’s nice and easy upfront costs to get that stuff going (it’s virtually nil), but the ongoing costs and delays are much higher.
Looking at your own business model, you’ve got to look at which one of those is going to work for you best.
Merchant fees is another area to look at. I think in Australia, we’re up to about 700,000+ small businesses that are selling online. Merchant fees are a real problem, and handling them in an accounting sense is a major issue. There aren’t many systems do that very well. We handle it pretty well now, so you can save a lot of labour by having transactions come in and letting the automated process deal with the fee. It books it to an expense account and everything.
Using Bank Automation for Expenses
|Bank Feed Support||Employee Time||Credit Risk||Cost Time/$|
|Debit Cards||X X X||X X X||X X||X X X|
|Credit Cards||X X||X X X||X||X X X|
|Employee Credit Card||X||X X X||X|
|Single Purchase||X X||X X X||X X|
|Petty Cash||X||X X X||X|
I see a lot of people use expense systems. This usually involves having their staff fill out forms or actually buying expensing systems. I’m personally not a big believer in those approaches. At Saasu we prefer to use debit cards. They provide some really good credit risk safety in the sense that you just load them with limited amounts of money. So that amount you load in is your maximum loss if the card is stolen, lost or whatever. There’s a bit of time in setting them up, but they’re excellent because you can get bank feed data from them and it just enables coding of transactions better. They also tend to be pretty well supported by fee providers. Credit cards are too, but not as well supported. Having an employee credit card is high risk. It still saves a lot of time, but the employee is taking a risk and it’s just not as efficient a way of doing things.The fourth on the list involves entering things in a single purchase or in an expense system. I just think that’s hard work. I think we’re past that now, and we need to move on from those techniques. Lastly, petty cash. This one I just think you should completely avoid if you can.
Transaction Workflow Automation
Create From a Sale (or Purchase)
Recurring Sales or Purchases
A lot of the people that use our system to do e-commerce asked for this feature. They really wanted to created a sale from a purchase. This is just a really simple thing that you can do. When you create a sale, you can actually create a purchase off it instantly just by using that ‘Do More’ option. There’s a recurring sale or purchase type format as well, for the subscription economy.
My observation is that people fear automated transactions a bit. The thought of sending out an invoice that hasn’t been checked, automatically, is justifiably scary. To overcome that fear, just get it set up and have them set up in a way so they don’t send out the emails automatically. That gives you a chance to check them, and it gives you a chance to work out where there might be problems in the automation. It allows you to be comfortable with it before anything gets sent out. That’s really how to get around that problem, and you really won’t look back when you do this. It’s how we run our partner network and it works really well. You’re always going to have the odd teething problem, but it’s way better than trying to deal with stuff manually.
Bank Rules to Automate Reconciliation
Bank rules are just a way to tell Saasu how to automatically assign a transaction to an account code. The way it does that is it looks at the description line in a bank transaction. It looks for keywords that you tell the system to look out for. It might be the word Telstra, for example, and when you see that word you might want Saasu to code that to an expense account for telephone. It’s really easy. Some accountants are really good at doing this stuff, obviously. I think 70%-90% of data entry can be done from bank feeds and bank rules, capturing and dealing with that data automatically. If you’re a micro or smaller small business, this is really useful. As you get bigger it becomes a bit tougher. To be honest, using straight through processing is a better technique.
Automating the Customer Experience
After Market Service
Customer experience is another area of automation to look at. There’s a lot of things that make up a customer’s value to the business: how often you get them to come back and buy from you, how much they talk about your brand, your ability to offer after market product and services. There’s lots of stuff: recurring revenue, up-selling etc. I just wanted to quickly explore some ideas around that. There’s bits of automation you can do inside this area.
Apple does customer experience really well. That photo is of a Genius Bar. Many people would be familiar with them. The purpose of these sort of services is not just for service. Apple sees this as a way to re-engage the customer, keep Apple at the front of the customer’s mind, and remind them why they love Apple. For Apple, it’s not really a cost base. They see this service as a sales channel, which I believe is an excellent model. So stop thinking of service as a cost.
It then begs the question, “what do you do to automate some of this stuff?”. Now, you’re always going to have humans involved in service, but there’s a lot of stuff that can be automated and still provides a great experience.
Managing Customers Asset Cycle
Warranty : Servicing : Refreshing
These are some ones that I’ve seen that I think are excellent. You can deal with things like warranties, service calls, or refreshing products you’ve sold to customers, be that reselling a new product for a product that ages etc. Attributes inside Saasu are really good for that.
Warranty, Leases, Servicing
Electronics E-tailer Serial Number, Colour, Memory Size, Connectivity Type, Warranty, Lease Health Food Shop Serial Number, Expiry, Batch Number, Recall date Music/Software Product Key, ISIN, Expiry/Renewal Publication Seller Renewal date, Author, Publisher, Format, Genre or Category
In the screenshot above, there’s a lease expiry and a warranty expiry set up as attributes with dates. The nice thing is you can then go and run reports for time periods and see things that are expiring, which enables the sales team or the service team to then go and contact those people and deal with those upcoming events.
You could also pass this kind of info across through an API to CRM systems, if you want to do things that way. Think of all the stuff that you could actually trigger automatically from having the right sort of data inside the system. Stuff like extending warranties, doing maintenance call outs, recalling unsold goods from consignment etc. There’s just dozens and dozens of uses for this type of thing. You can do it inside a product like Saasu or inside products like Sugar and Salesforce and so on.
Building Back Office Systems
We’ve got a back office system that runs behind the scenes in Saasu. It’s actually just an extension of the manage subscription area that customers use. Ideally, we have this eat our own dog food policy. We use our own system, we live and breathe it so we know what’s wrong with it as well as the pros and cons of it. We also try to build stuff that we can give to you guys down the track. If you’re in the software game, I’m a bit of a believer in building systems attached to your core offering if you can do that. If you’re in a service industry or building physical products, that might be a bit different.
7 Rules of Stunning Spreadsheets
I kind of want to remove the bad name spreadsheets get. A lot of our competitors out there are kind of dissing the spreadsheet, but I still think spreadsheets are excellent. Don’t be ashamed if you like them too! They’re still unbeatable in so many ways, especially with extensions like Pivot in the screenshot there. I think Microsoft owns it now, but it’s by Live Labs. You can visually display information that’s coming out of spreadsheet data in a really nice way.
There’s lots of cool things out there you can do. You can connect spreadsheets to APIs and do all kinds of things. It does take some technical skill, but I still believe that if you’re trying to automate tasks inside a business, you can do a lot of that with clever spreadsheeting. It’s also a lot cheaper sometimes than trying to go out and hire programmers to do that stuff. Take a look at my seven rules for stunning spreadsheets below for a couple of tips on how to use spreadsheeting to achieve automation tasks inside a business.
- Version control.
- Less spreadsheets, more tabs.
- Learn Macro’s and XML interaction.
- Learn about tools like Pivot that work with spreadsheets.
- Don’t listen to people who say don’t use spreadsheets.
- Spreadsheets must have a test cycle.
- Spreadsheet manipulation must be limited. Lock fields & formula’s
Buying Automation Tools Off The Shelf
You can look into using external tools, buying automation things off the shelf like OneSaas. There’s a bunch of data brokers out there you can use, and connecting systems with off the shelf tools. It’s usually only a few easy steps. Some people think it’s hard, but if you just follow the link from OneSaas on how to set up a connection to Saasu to MailChimp or Campaign Monitor, there’s literally six steps. It’s super simple. It’s going to take less than five minutes to get it up and running. Just think of the mental roadblock versus the actual time cost in doing that. Get over the mental roadblock, try it out and see if it works for what you need. You might get a huge ongoing pick up out of it.
Outsourcing multiple tasks, knowledge and complexity is a common mistake.
Single tasks outsourced reduce execution, security, dependency & quality risk.
The whole Mechanical Turk thing came from this fake chess-playing machine in the late 18th century. A person would hide inside this box and operate the machine, giving the illusion that this ‘machine’ could beat everyone at chess. Fake mechanical automation essentially. Guys like Amazon have taken that on board, and have this team of Mechanical Turks you can hire. Fiverr and Freelancer, you could regard as Mechanical Turks. It’s using humans to do tasks. They call them Mechanical Turk workers. Amazon actually pitch it as artificial artificial intelligence, which is quite clever. Mechanical Turks is definitely another option. Some of the most successful businesses we’ve seen on Saasu have almost unanimously been involved in outsourcing to these sort of services in some way, simply because they’re able to get so much bang for their buck.
This sounds to good to be true. Is this really possible? Yes.
The subscription economy revolution
In the financial accounting space, I think one of the key automation areas is looking at how you price and bill as a model. A single one-time transaction is how the world traditionally worked. To be honest, if you are paying attention to the stock market you’ll find out that those business models aren’t worth much anymore. You might be lucky to get one to two times revenue for a business that works that way. There’s this huge equity value pressure to get away from that model.
In the ’80s and ’90s, everyone was sort of doing user or unit pricing, that is that software was sold per user, or services were often related to a unit or a time box amount of time, like a call out fee for 60 minutes worth of value, things like that.
Then a shift into usage based billing happened. The telcos are obviously big in that space.
But today, it’s really the subscription economy that’s evolving. The highest value companies in the market now tend to be ones that have this recurring fixed or variable revenue type models, because that’s repeat revenue. Obviously the valuations go up, but the nice thing about that is it’s really open for automation. So you have this trickle incentive, it’s really good for cash flow and it reduces cash flow volatility. The customers like it because they’ve got certainty about where they’re spending. It’s also very automatable. It’s definitely something worth looking at within your own business model to see if it applies to you.
One Time :: Unit Pricing :: Usage billing :: Recurring
Automating debt collection using statements
Here’s something we obviously use ourselves, statements inside Saasu. We use them just for reminders for our accountants. The accountants listening will know it can be a little bit annoying getting a reminder, but I have to be upfront with you and say I really don’t want to be having a negative conversation with our partners, who I really appreciate, about collecting money. I’m glad a system does it and I don’t have to get involved. I really believe in letting the system be the bad guy in this situation and just building a nice build up of emails.
I think we have three templates. We have one that goes out just as a reminder when their account is due, one when it’s really overdue, and then a final reminder or final warning type one. We’re not regressive on AR as some companies are, but you do find that there’s companies that need to send legal letters. You can put that template inside Saasu. You can have it coming from your lawyer with their approval, so it goes out automatically if you’re trying to collect money. So there’s ways to use Saasu to automate all of those tasks.
Automating Brand Management
Brand guidelines help automate design
Automating design and brand guidelines is really a critical thing to do. This is the automation of how you present yourself to the market at the end of the day. Brand guidelines are an excellent way to make it really easy for people in your organisation to know how to present the brand. That might be in how they design a brochure, do a presentation, or how they build a webpage. Lots of things come up with branding.
You really don’t want everyone to have to keep reinventing designs. If you have brand guidelines built, you’re really giving people the explanation as how to go and design something as best they can within their ability. I still think it’s worth engaging designers to make sure people are adhering to the brand, but also to refine and update it over time.
This doesn’t just apply to us as a software company, it also applies equally to small businesses. You’ve got content, collateral, websites, and display materials you need. All of that stuff has to be managed. If you have brand guidelines, it’s just so much easier to have that happen. You can give the brand guideline to someone who’s helping to do those jobs, and it’s really clear for them what things need to look like, otherwise you’re wasting big dollars every time reinventing the wheel.
Pattern Libraries Help Automate UI & UX Design
Pattern libraries are another way to help with that. That’s really about how things should look inside an application, but it also applies to other areas. Things need to look a certain way so they’re on brand. There’s also certain ways elements work, such as a certain way you might use a particular type of drop down in a website, or the type of button you use and how it works, things like that. Pattern libraries can be good if you’re a more tech related business. Twitter and Yahoo! have really good pattern libraries that are public. You can check them out if you want to get a feel for what they are and how they work.
Automating quality assurance
Standard Operating Procedures (SOP’s)
I really believe in this as an area of automation. It mightn’t seem like automation at first glance, but standard operating procedures allow you to work out what you need to automate next. They’re like a feeder for automation. You’ll find stuff in operating procedures that are really long and tedious and you’ll recognise that that stuff is chewing up a lot of time. If it’s taking 20 steps to do this, is there a better way? Standard operating procedures almost create a specification for designing the things you need to automate inside the business.
Read E-Myth if you haven’t read it. It’s an old book now, but it still totally applies. Having a really decent operating manual is effective. The image above is inside a space shuttle, and that’s a paper based SOP document they have. Despite all the space shuttle’s computing power and tech, they’re still using paper and it’s effective. It’s because they’re trained in rituals and doing things a certain way as quickly as they can. That’s sort of the automation part of their role.
Debugging your product or service
This isn’t just about debugging software, this is about what your service offering is. It’s about the things you manufacture and how you get that working really well, and how you create quality. In our business, we write a lot of test scenarios every time we write some code. It reveals bugs as people are writing the code too, so it catches things nice and early. I think it applies to manufactured products and services too. You should be looking at having people review what you offer or sell, and getting feedback on that. Test things early and often before you go to market fully with your products.
I think we’ve got about 5,000 automated tests in our run over the Saasu application. They’re running every day as people add code to the database and build it out. That’s a massive reduction in labor when you think about the implications of that. Even when you’re building systems in small business you can still automate a lot of tests yourself.
A classic example is building what’s called a checksum inside of a spreadsheet. It just checks a couple of different columns to make sure they add up to zero or some number, and then you know the spreadsheet’s working. You might get a little bit of code written that goes and checks your e-commerce site for the total sales today and compares that to the total sales in Saasu to make sure everything’s running smoothly. A little script like that would be cheap to build. Maybe you could have self balancing tests, looking at the total debits and credits in your accounting file which should add up to zero. We have a simple test like that just to check. That’s a quick sanity check of a file for example. There’s lots of ways you can do this.
The secret to the highest quality organisations in the world is Checklists
If you’re not technical and you’re needing to do manual automation, having checklists is another great way to automate. Think of it like the classic pre-flight check on planes before take off. They’re just to make sure everything’s in order. It’s not just about that though. What they’re doing is they’re automating the process so it happens really quickly and systematically. People get really fast at doing rituals and the checklist helps to facilitate that. Checklists just create efficiency, accuracy, and speed, essentially. Once you’ve got a checklist set up, you could then look at ways to have systems do the checking for you. Things like sensors, feedback mechanisms or beacons are great for that.
Sensors, Feedback, Surveillance and Beacons
There’s hundreds of devices coming out now in the Internet of Things, which is really just objects that are connected to the internet. These connected objects can give you feedback. Say you’re hiring cement mixers, for example. You might put one of these little objects on your cement mixer. You know where it is, you know where it’s been, you’ve got ways of tracking the mixer and you’re comfortable that it’s being looked after in a very automated way. You can have trigger events too if something’s gone wrong, say if someone’s driven it up to Queensland and you’re not going to get it back again. Cameras are a good example of this. There’s lots of different ways you can use sensors and feedback to try to automate sort of the quality side of the business or the monitoring side of the business.
Automating growth monitoring
So the last section is monitoring. Obviously, I’m going to talk about the Saasu dashboard. That’s the thing I use the most in our business, probably just as much as I use Geckoboard. Dashboards are a facilitator for accurate thinking. I love that term. Sofia, our previous COO, introduced that to me. They help you to get a clearer picture of your situation and identify gaps. You can do a proper assessment of the data you’re shown, then proceed to verify and design your actions around that factual evidence.
If you’re a small business you can waste a lot of time trying to build a holy grail dashboard. I think business is too complex for that. You’ve kind of got to find off the shelf dashboards that work well. Spend your time using those. You’ll get plenty of data off them. You’ll see lots of holy grail type dashboards coming out of the market, but it’s almost too much on one screen. You’re not going to be able to get it all. You do need to have dashboards from different systems.
Salesforce is another system we use, and there’s great dashboards inside. The object is to really get as much data through as possible without having to generate that yourself. You want all of that data flowing through in an automated way. The second step is really curating down to the things within that data that are really important. You’re looking at backward-looking data or KPIs, present day type KPIs, and forward-looking KPIs, in each key area of the business. If your dashboards are covering those KPIs, then you’ve got enough information to be making sure you’re making the right decisions.
As you know, in business we’re all about selling emotion and experience. I’ve been talking about that in quite a few of these presentations. It’s not just about selling products or services. Data is a way for us to enter this world of informatics, which is really about building systems that deal with the psychological and behavioural things that customers do. You need the data as the first step to that.
Marketing Dashboard & Analytics
This is an actual screen in our Mixpanel. I’ve wiped the numbers though, just for a bit of privacy on our data. To give you a bit of an idea, this is just tracking things like notifications we send out, whether people are clicking on help links, adding bank accounts etc. We can kind of monitor the success rate of those kinds of activities in a consolidated way to make sure the product has been adopted and used the right way. Then we can have trigger events off the back of that.
For example, if someone was failing to connect a bank account. One of the things we want to have set up is we want to be emailing them and offering them a walkthrough or hand hold through that process. Or if they’re trying to migrate data and they get stuck, systems like this are useful in helping with that. They can automatically fire an email off to a service person who can then call the customer and help them out. This applies to any business model, services or products. In this example, one of the things that might happen is about seven days after someone signs up to Saasu and they haven’t been active, we can fire off an email offering some assistance. Even if they haven’t been active for 30 days, we might be able to offer them a trial extension.
The last one we’ll talk about, the KPI dashboard, is really important. This is Geckoboard. Again, this isn’t our data, but it kind of gives you a bit of a feel for what it looks like. The idea is that you have all your data at hand from different systems. It’s what I was talking about before, a conglomerated view of the world, which is very useful. Sofia, our previous COO, used to say, “if there’s no scoreboard, you’re not playing the game.” That’s so true. If you don’t have one of these in your business, you haven’t got the scoreboard, how are you going to achieve the goal? How do you know how you’re going versus the goal, timelines, and things like that?
This product, Geckoboard, is pretty easy to set up. There are lots of these products out there now, but this is one of the better ones. It connects mostly to CRM systems and email systems. That sales and marketing cycle is really the focus for hitting your revenue, so it’s probably a good area to start. If you’re using other common systems like Google, for example, it’s really easy to connect to Geckoboard and get data flowing down into it.
These are systems that the bookkeeper, advisor or the business owner can be tapping into. They’re way more detailed, and pay more respect to the business owner than a lot of dumbed down systems which might just look at a few key numbers and don’t really get into the nitty-gritty. I think we’re getting sophisticated in business, and advisors and bookkeepers have this advisory capacity they need to step into. A lot of our partners are already like this, but the good ones are really stepping into the detail, the nitty-gritty, inside a business. That’s where they’re adding the value and obviously able to charge for it.
Informatics looks at interaction between humans and data. At Saasu we have our data highly visible. In this screenshot, you can see the screens in the background. What you’re seeing there is the informatics part of it. This is us looking at our data and going, “what’s our market? What does the data imply about the psychology of our market and how they’re operating?”. It’s also about having a data culture. We need data to be highly visible to do that, not just in a mental way, but with the physicality of having it on screens.
To be honest we walk the talk. We didn’t walk the talk for a long time, but we’ve really got it all set up now. In those screens, we’ve got Team Foundation Server, Geckoboard, one screen is reserved for news, and another a Salesforce dashboard. We’ve got good coverage of what’s happening in the business.
Automation Implementation Checklist
Story photo by Alex Knight