d78c0df3

How to carry out a content audit

Wed, 23 Dec 2020 17:29:02 GMT

A content audit, not be confused with a content inventory, which is simply a detailed list of all the content that you have, is an in depth assessment and evaluation of all your website content to determine its relative strength and weakness in relation your content marketing strategy.

An audit will help you identify

Ways to improve organic search performance.
Which content pieces perform well and which do not.
Content gaps.
Opportunities to repurpose existing content.
New ideas for future content pieces.
Content that is no longer congruent with your business.
Ways to improve the information structure, i.e. content accessibility.

To help you conduct a content audit, we have outlined a few essential steps.

Step 1 – Create a content inventory

You first need to know what content you have before you can decide what to do with it. Start by creating a spreadsheet where you can build out a complete picture for each piece of content. Details that you might want to consider capturing for each content piece include:

URL
Page Title
Target Keyword
Audience (who was this content intended for)
Meta Description
Page Headings
Links
Image ALT Tags
Date Last Updated
Page Visits (based on a 3 month average)
Page Bounce Rate
Average Time on Page
Word Count
Type of Content (article, blog post, informational page, landing page, infographic, etc)
Content state (current, out-of-date, Evergreen)
Content Owner
Call to Action
Conversion Data

Step 2 – Analyse and identify what actions are needed

Armed with this wealth of data you now need to evaluate and agree what actions are required for each piece of content. When evaluating your content look at it from the following perspective:

Is the content attracting the right people?
Is the content supporting the desired outcome?
Is the content engaging?

Having evaluated your content against the above, the next step is to decide what actions are required. Typically this would be:

Keep best performing content.
Edit content that performs well(ish) but could do with a refresh.
Delete content that does not serve your goal or your user community.
Consolidate content that on its own is weak, but consolidated will produce a much stronger piece.
Reorganise content to a place that makes it more accessible.
Create new content required based on the content gaps identified.

Step 3 - Create a plan

Next, assign owners, determine priorities and set deadlines so that the actions that you have identified above are carried out in a timely fashion. Deadlines need to be realistic and it is always a good practise that when assigning priorities you look for the quick wins first.

A content audit is a valuable tool to help businesses better understand what content they have on their website, the value that this content is currently adding and what needs to change to better serve the organisational objectives and user community.

An overview to software testing

Wed, 23 Dec 2020 17:28:59 GMT

When testing a new website or piece of software there are various tests that you can undertake to ensure that the end product is bug free and of the highest standard.

Before delving further into the different testing types it is important to note that the level and amount of testing that you undertake is largely dependent on the business requirements, complexity and size of the project.

The testing that we will be looking at includes:

Usability testing
Functional testing
Compatibility testing
Database testing
Security testing
Performance testing

Usability testing

Usability testing is nothing more than assessing how easy your product is to use and whether the flow through the solution is logical and intuitive.

Usability test scenarios include:

Correctness of content.
Use of typography.
Usefulness of error messages and tooltips.
Flow of content through the product.

Functional testing

Functional testing is testing to see whether the features that have been developed perform as outlined in the specification and tests all end-to-end workflows / business scenarios (both positive and negative scenarios).

Functional test scenarios include:

All mandatory fields are validated.
Leap years are validated correctly and do not cause errors and or miscalculations.
Numeric fields do not accept non numerical inputs.
Max length of every field to ensure the data is not truncated.
Correct currency is applied and that calculations work.
Any functional failures deliver meaning error messages.
Uploaded documents are opened properly.
Email notifications are sent for different transactions.
Forms are working as expected.
All links are working (outgoing, internal, anchor and MailTo links).

Compatibility testing

Compatibility testing focuses on making sure that your solution is compatible with other elements of a system with which it should operate, e.g. Browsers, Operating Systems, or hardware.

Compatibility test scenarios include:

How does the solution display in different browsers (IE, Firefox, Chrome, Safari and Opera).
Is the HTML version compatible with browser versions.
Do images display correctly in different browsers.
Are fonts usable in different browsers.

Database testing

Database testing looks at whether the records inserted in the web application are stored correctly in the database.

Database test scenarios include:

Tables, columns, column types and defaults match the specification.
Primary and foreign keys of each table are correct.
Stored procedures work correctly.
Data is saved correctly in the database after each page submission.
Response time of each query executed.
Data displayed on the front end is the same as the back end.

Security testing

Security testing is undertaken to identify any flaws and gaps from a security point of view and include:

Important data is encrypted and submitted via HTTPS.
Password rules are implemented on all authentication pages.
If the password is changed the user is unable to login with the old password.
If the user is logged out from the system or the user session has expired, the user is unable to navigate the site.
Can secured content be accessed without login?
User account is locked if the user enters the wrong password several times.
User roles and rights have been correctly configured.

Performance testing

Performance testing is conducted to evaluate the compliance of a system or component with specified performance requirements and covers.

Performance, stability and scalability under different load conditions.
Can the architecture support the solution at peak user levels?
User response times.
Stress testing the product to determine its breakpoint.

Testing is an essential part of the development lifecycle and needs to be undertaken at multiple points to ensure a high quality end product. The investment in time is significant and when building out your project schedule make sure that you have allocated sufficient time to this crucial phase of the project.

Why projects fail

Wed, 23 Dec 2020 16:59:42 GMT

With more than 20 years project management experience under my belt I have seen it all – hugely successful projects and colossal failures. Interestingly though, the reasons why a project succeeds or fails is the same, regardless of its size or complexity. So, when looking to embark on a new project bear the following in mind:

Insist on the truth – always!

You will be amazed how often the truth is slightly bent during the lifecycle of project, for example, the project is running late, but instead of being told that the project will not be delivered on time, the supplier will say something like, “at this point in time we are slightly behind schedule, but we will still complete the overall project on time”. Anything that starts will, “there is a slight issue, but . . . .” is a red flag and you would be remiss to just accept that. Interrogate the statement to understand the full impact of the issue and find out what remedial action has been put in place to mitigate the risk and protect the delivery.

Lowest price

You’ve heard the expression “buyer beware”, well this is very much the case when undertaking a project. Be cautious of the supplier who comes in with the lowest price. I do not want to paint everyone with the same brush, but from my experience if a price seems too good to be true, it usually is the case! Some suppliers will undercut their prices significantly just to get your business and then, during the execution of the project you are given one “change request” after another. The best way to protect yourself against this is to get the supplier to outline in detail exactly what they will and will not deliver as part of their quotation, i.e what is in scope and what is out of scope.

Not assigning the correct level of attention or people

An important factor to the success of the project is the project team. You need people on the project who are experienced, can provide valuable support to the project and, more importantly, are able to invest time in the project. Assigning the project to the latest recruit, someone who is already over extended, or someone who does not know enough about what problem the project is trying to resolve, is a recipe for disaster. If you do not have suitably qualified resources for each phase of the project it is in your best interest to bring those skills on board to assist.

Lack of communication

Very similar to the point above, communication throughout the project lifecycle is crucial. You need regular feedback from your supplier, your project manager and the team delivering so that you are not caught unawares and can plan appropriately for the delivery. If the project will help you better serve your customers, improve a process, resolve an issue or assist in generating more business, you want to know how it is progressing. They say that forewarned is forearmed and this is especially true for projects – with regular communication you can plan and deal with issues far more proactively. At the outset of the project agree a communication plan that clearly outlines what will be communicated to you and at what intervals.

Lack of planning

I am sure you have come across the 6 Ps, i.e. Proper Planning Prevents Piss Poor Performance. This is the cornerstone to a successful project and will help address all the points above. You need to be planning before the project even starts, for example, what does the project need to deliver, what evaluation criteria will you use to select the best solution, when is the best time to undertake the project and do you have the necessary resources (budget, people and time) to undertake the project? Once the supplier has been selected, more planning is required. This is mostly done in conjunction with the supplier and includes assigning the best person to the project, understanding what will be delivered and by when, knowing what will be communicated to you and understanding exactly what is expected from you during the project. And just when you think that you are done you need to plan for when the project is actually delivered and how this will impact on your customers, your processes and staff. You need to plan throughout to give the project the best chance of success.

Do not take any project for granted, even the smallest project, if delivered badly, can have a significant impact on your business. Invest the time and effort into planning.

Remote/virtual team businesses – should I use them?

Wed, 23 Dec 2020 16:57:15 GMT

When I wrote this original blog post back in 2017 about the benefits of using the services of a virtual or remote business, who would have believed that we would be hit by a global pandemic that would change the working landscape forever? But here we are in 2020 where we have seen that most businesses had to implement remote working as a matter of course. The pandemic has opened our minds to remote working and shown that my original comment is true, i.e. it is not the brick and mortar structure that defines us but rather our capabilities, skills and expertise. As a result of Covid many traditional businesses have shown a willingness and commitment to continue with remote working in some way, shape or form and I think that this presents many opportunities and benefits to companies and their clients that they serve.

It is also worth noting though that as a result of Covid, and the many restrictions that were imposed, the companies that I had originally listed below as the top virtual companies are no longer the largest virtual companies. That being said they are definitely the pioneers of this movement and already had robust processes and systems in place to continue to function uninterrupted when the pandemic struck. Many organisations “forced” into remote working on the other hand had to make significant adjustments.

Covid has made remote working a reality and I would hope to see that all companies that have adjusted to and plan to continue with remote working pass on the savings and efficiencies to their clients in the same way as those remote companies who have been virtual from day one.

Original post:

The traditional office environment has changed. Gone are the days of working 9 to 5. Flexi-hours, work from home and geographically dispersed teams have become the norm. Actually, businesses are now even trading brick and mortar buildings to become full-fledged virtual businesses. The biggest virtual companies to date are:

Scopic Software
Articulate
10up.com
Clevertech
Basecamp
Buffer
X-Team
Scrapinghub
Zapier
Collage.com
Fire Engine RED

So, should you use the services of virtual companies? Without a shadow of a doubt. The benefits that companies derive from going virtual are passed on directly to you, for example:

Lower operational costs

This is a no-brainer. No office rent, significant savings in office furniture and décor are just some examples of overheads that are not incurred and thus frees up funds. These funds are in turn used to attract highly qualified staff, are used for innovation or service refinement and, more importantly, the savings are passed on to you, giving you a greater return on your investment.

Office hours

When working in a traditional office environment you have set working hours, and more often than not, employees became prisoners to those office hours. If your office hours are 9 – 5, how often have you come out of your office or a meeting just past 5 to find that most of your colleagues have left? Interestingly research has shown that virtual staff tend to put in more hours than expected. People get in “the zone” and forget about time and you find that more is achieved and delivered.

Access to a larger talent pool

One of the biggest challenges that remote working offers is access to great talent. The location of your business invariably reduces the number of people who will apply for a role. Going virtual - problem solved! Virtual businesses have access to an international resource pool meaning that you no longer have the restriction of time zones or locations and you could have access to the best of the best and not have to settle for what is available.

Greater productivity and value add

Yes, this is true. Virtual teams are happier and more productive because they are better able to manage their hours and, more importantly, are not distracted by office politics or the office clown. The virtual team gets on with the job.

We are living in a time that is continually evolving and with this evolution comes new ways of working - we should no longer be defined by a physical location, we should be defined by what we do and the value that we add.

What is the difference between a web app and a website?

Wed, 23 Dec 2020 16:47:52 GMT

While the term “website” is generally used for both websites and web apps there is a distinction between the two that is worth noting.

To explain this distinction, let us compare 2 online profiles, i.e. a digital marketing agency that “sells” digital marketing services like SEO and PPC and a company selling office furniture.

The digital marketing agency provides the user with information about their service offering, an introduction to their team, contact information and some client testimonials. If the user wishes to purchase any of these services, they would contact or meet with the Agency, request a proposal and if happy with the terms, they may agree to contract.

In this example we find that the content is purely informational, there is very limited interactivity with the content and the content is mostly static. These characteristics are typically what defines a website.

Now let’s take a look at the office furniture seller. In this example you can learn more about the product, find out about the company, and place an order. As part of the order placement process you can customise your product to suit your requirements, create an account and checkout. As a user you are interacting with the solution, manipulating the content and getting a result that is unique to you based on the parameters that you have entered. The level of interactivity and bi-directional flow of data is what makes this more of a web app as opposed to a website.

So to summarise, a website:

Is primarily created for the purposes of providing users with information.
There is limited, or no interaction with the web pages. Interactions typical for a website include newsletter signups, contact form submissions or search.
Presents the user with digital content in the form of images, video, and audio.
Primarily static, meaning every user sees the same content.

A web application on the other hand:

Is a solution that is built to provide an interactive experience specific to the individual using it.
Allows users to interact and manipulate the data and presents different information based on this interaction.

The above differences are by no means exhaustive as there are other differences as well that I have not covered, for example:

The function of a web app is a lot more complex than a website.
Integrations for web apps is a lot more complex because of the richer feature set available.
A web app typically requires some form of authentication to allow the user to perform certain tasks.

While there are many factors that distinguish a website from a webs app, the primary distinction between the two is user intent, i.e. a website user is focussed on consuming information whereas a web app user is primarily interested in interacting with the system.

How to report a bug

Wed, 23 Dec 2020 16:40:27 GMT

You are testing your website and find that when trying to make an online purchase you cannot progress beyond a certain step, or while trying to sign up for an event you find that you cannot complete the registration process due to a technical issue. What do you do? Well, you fill in a bug report of course so that the developers can investigate and resolve. This is the normal process and below is an example of what a “typical” bug report looks like.

“I was trying to checkout on the site and got an error message when I clicked submit. Please investigate.”

Does this look familiar?

The problem with the above is that while the information provided in the issue report is factually correct, there is insufficient information for the developers to investigate. This is a common problem caused by the fact that there is generally a disconnect between the information that the user provides versus the information that the developers want when receiving a bug report.

So, what exactly are developers looking for when reviewing a bug report? As much information and context as possible is the short answer. Without this you find that there is a significant amount of time wasted toing and froing so that the developer can better understand the steps you took that resulted in the error. You need to remember that every feature on your website is essentially hundreds of lines of code that the developer would need to review to try and determine what is causing the issue. The more information you provide the easier it is for the developer to home in on the most relevant piece of code.

So, what does a good bug report look like then? If we stay with the example listed above, a good

version of that same bug report would be:

“While trying to check out the following product (URL to product included) I encountered the attached error message (screenshot attached) on screen 2 of the checkout process. I filled in my name, billing and delivery address and contact details and then clicked the next button.
I added the product directly from the product page (not from the search results page or my wishlist)
I am running windows 10 and the browser that I am using is Chrome. Chrome is currently running the latest version (87)”

Side note - to find your browser and operating system you can use something like https://www.whatismybrowser.com/

Notice the difference in the level of detail provided. The developer now has all the information that he needs to investigate further or follow the same steps that you executed to try and replicate the issue.

The problem with the above is that unless you test websites regularly you do not naturally think this way. Most users typically focus on the end result and not the steps that they took to get there, which makes it very difficult to provide the level of detail that the developer ideally needs to troubleshoot.

The solution? Outsource your end to end testing to someone who is trained to actively look for bugs and who can provide the right level of detail needed by developers to restore the feature to a working state in the shortest amount of time.

Why content migration can be difficult

Wed, 23 Dec 2020 16:31:20 GMT

When we think of content migration, we think of the physical task of moving the content from one CMS to another. This is, however, only one aspect of content migration and while exceptionally resource intensive, is by far the least complicated.

The more challenging part of content migration happens long before the act of physically moving the content and is more of an organisational issue than a technical one.

Content creators can be “emotionally” attached to the content that they have created and see it as a validation of their value and work. Notifying them that their content has not made the cut can, as “silly” as it may sound, cause them undue anxiety and unhappiness. These conversations need to be approached diplomatically and one way of avoiding a situation like this now and in the future is to introduce a content expiration program whereby you assign expiration dates to different content types. If content contributors are aware of this practice, they do not become emotionally attached to their content allowing you greater control over which content to “expire” on the site if need be.
When the size and scope of the content migration effort is significant there is a tendency to “migrate as is” on the understanding that the content will be reviewed at a later stage. In all my years of working with content this never happens – it is a pipe dream. If you don’t have time to review the content prior to launch, why would you think that you will have time to review it after launch? Start early with your planning so that you have a clear path to migration, from the initial content audit right through to content upload and sign-off. Identify who will be responsible for the migration, who will deliver what content and who is ultimately responsible for signing the different content off.
Many people believe that content migration is a simple “cut and paste” exercise so do not see the need to invest time thinking about it upfront. If this is the mindset that you are facing you have an uphill battle ahead of you plagued with little or no commitment or a sense of urgency from content contributors. The best way to deal with this is to make sure your senior management team are fully aware of the significance of this phase of the project and have nominated a content ambassador who can drive from the top down and support you when you encounter any resistance within the organisation.

The above organisational challenges can turn what is already a difficult undertaking into a very stressful event and place a lot of unnecessary strain on a team. It is for this reason that many organisations insist that the content migration is managed by the agency building the new website or an outsource agency who have experience dealing with the difficulties inherent in content migration.

How to plan for content migration

Wed, 23 Dec 2020 16:23:41 GMT

When building your new website, an important consideration is the migration of the content from the old site to the new. Content migration can be both a significant and difficult task and needs to be considered right at the outset. Leaving this to the last minute has been the single biggest contributor to delayed site launches.

To help you plan your content migration we have put together the below guide which, while not exhaustive, will definitely start the process on the right footing.

Step 1 - audit your existing content

Before you migrate any content you first need to audit the existing website in order to understand exactly what content is on the site. Once you have determined what content is on the site, you need to ask the following question for each piece of content:

Is the content still relevant?
Does the content need to be edited or can it be moved across as is?
Does the content warrant its own post, or can it be merged with another piece of content?
Should the content be deleted?

To find out exactly how to carry out a Content Audit, read the following blog post.

Step 2 – map the content to the new information architecture.

Once you have determined which content needs to be migrated across, the next thing to consider is where will this content reside on the new site. More often than not when you design a new site, the structure, or information architecture, will differ from the original site. Mapping where the content will go in advance makes the eventual migration smoother and allows you to identify content gaps early in the process.

When mapping the data it is helpful to create a spreadsheet that contains all the relevant information associated with that piece of content, for example current URL, where it must be placed on the new site, meta description, H1 title, internal linking, etc. The more information you add to the mapping spreadsheet the easier the migration process.

Step 3 – Decide on an approach

It is best to decide upfront how you plan to approach the content migration process, i.e.

Migrate the pages in terms of popularity - pages with the highest views are moved across first and those with the lowest views last?
Migrate by section - blog content followed by resources followed by key landing pages and then everything else?
Move the content across as is and revisit later to add revisions and metadata, or make the changes while you move the copy across?

Step 4 – Migrate the content

This step is self-explanatory and involves physically moving the content across from the old site to the new site. When moving text, it is advisable to first drop the text into notepad before adding it to the new CMS. The reason for this is that when you put the text into notepad, notepad removes all styling and formats which means that the text will not display incorrectly because styles that are not compatible with the new CMS have been copied across.

Step 5 - Review and Signoff

Once the content has been moved across it is advisable to get the different content owners in the organisation to review the uploaded content in terms of completeness. This should also include a device and browser review to make sure that there are no inconsistencies in how the content is displayed to different users.

A final point to note is that when you put together your migration plan you must also put a content freeze in place. The content freeze ensures that no content is added to the current site after a certain date so that you know that at the end of the migration phase you have brought all content across and nothing has been missed.

Content migration is a very labour-intensive process that causes lots of frustration and unnecessary delays when done wrong. If there is even a slight possibility that you will not have sufficient time or resources to dedicate to this task then you should seriously consider outsourcing the work so that the launch date of your site is not left to chance.

Functional testing at a glance

Tue, 22 Dec 2020 22:34:42 GMT

Functional testing is essentially testing to make sure that the application or system is behaving as expected.

There are different types of functional testing and the amount and type of functional testing that you would perform is largely dependent on the type of solution, complexity and scale of the project.

Testing a brochure site for example would not require the same level of testing as an enterprise wide workflow management system. Broadly speaking though, functional testing can include the following:

Unit Testing – unit testing is usually performed by a developer and entails writing unit tests which call the methods in each unit and validates that when the required parameters are passed the value returned is as expected.
Smoke Testing - Testing that is done after each build is released to ensure build stability.
Sanity Testing - Testing that is done to ensure that all the major features of the application or system are working correctly.
Regression Tests - Testing performed to ensure that adding new code, enhancements or the fixing of bugs does not break the existing functionality or destabilises the application or system.
Integration Tests – Where a solution relies on any form of integration with another system to complete the feature set or pass data, it is important to test that this works perfectly together.
Usability or User Acceptance Testing - The final solution is handed to the actual customer in a production-like environment so that they can test the product to make sure that it performs as expected.

When undertaking any form of functional testing, you would typically identify the different test scenarios that you would like to test and use this as the basis against which the correct-and-completeness of the solution is measured. These test scenarios are known as test cases and detail the following:

Test summary
Pre-requisites
Test Steps and
Expected results

The more detailed the test cases, the more comprehensive the testing will be. That being said it is worth noting that attempting to write a test case for every scenario is sometimes not possible as it can be a very time-consuming and expensive exercise. It is therefore important to identify the most important and relevant scenarios that need to be completed at a minimum for the solution to be deemed deployment ready.

Functional testing is an important part of the product lifecycle and should not be ignored or made light of. Testing the solution using one, two or a combination of the testing types listed above as well as documenting the test cases and the expected outcomes is essential to ensuring that you deliver a solution that meets your client’s expectations. Not only will you have a client who is willing to sign the solution off, but you would have proactively dealt with any issues that could have caused a system failure post launch, resulting in reputational damage for you and your client.

While functional testing is in itself a comprehensive test type it does not cover the entire scope of system testing which you can read about in our blog post related to software testing .

Why it makes sense to outsource testing

Tue, 22 Dec 2020 22:34:42 GMT

When it comes to IT projects, specifically web development, many agency managers and project owners fall into the trap of believing that they have the necessary skills in-house, and more than enough time to thoroughly test a website before it goes live. The reality though, is the following:

The two phases of a project that are grossly underestimated, that cause frustration for both the client and the agency and that can impact your project schedule, are content population and feature testing.
Managing content population in-house is neither cost-effective or a good use of your team’s time.
Unless you have a dedicated testing team, the amount of functional testing done is usually inadequate to identify all issues before launch.
Testing is usually carried out by team members who are either too close to the project or do not have the requisite expertise to do the testing justice.
Projects rarely run to schedule due to unforeseen challenges associated with features taking longer than expected.

I am not saying that the above is always the case but based on my many years of experience managing web development teams this happens a lot more regularly than we would like to admit. The challenge with the above is that developers, PMs and Project Coordinators are so familiar with the project that when they do test, they use workarounds that “appear normal” to them and do not see the project the same way a fresh-eyed end user would. You will also find that they have other competing priorities which means that they run through the test cycles as quickly as possible so that they can get on with their other pressing tasks. This situation is further compounded if the project falls behind schedule in that the priority is all about bringing the project back on track and as a consequence other phases (including testing) suffer.

As a rule, it is never a good practise to have developers conduct end user testing as they are too close to the detail and truth be told they should be focussing on developing new code and refactoring or improving the existing code. The same can be said for PMs and Project coordinators. Their time is better spent on keeping the client informed, managing their expectations, managing scope creep, keeping a close eye on the budget and schedule, or taking on another project.

By outsourcing testing you immediately free up your staff to focus on the tasks that they are good at and that are more profitable for the Agency. The testing is carried out by users who have not been part of the requirements gathering, design and development phases and as such they have no bias towards the end product. They are looking at the solution in the cold light of day and will more readily be able to identify issues with functionality and usability than your own team.

Another challenge that outsourcing reduces is the quality of issue reporting. When you use your own internal staff, you find that the feedback is usually unstructured and inconsistent in the way that it is communicated and documented. The main reason for this is the informal nature of the relationships that exists with colleagues. External testers will have clearly defined processes and a standardised approach to feeding back on any issue so that developers can easily understand and recreate it if necessary.

There are many benefits of outsourcing some of your back office functions to other agencies that we outline in an earlier blog post , but looking at it purely from a website testing perspective, you will find that is surprisingly inexpensive, it allows you to redeploy your resources to activities that add more value to the Agency and client, and helps to identify any functional and or usability issues before being exposed to a “live” audience. There is nothing worse than handing over a project to a client or going live and then spending valuable (non-billable) time resolving a host of bugs.

In a nutshell, outsourcing your testing:

Means that your team can focus on improving the product and customer experience.
Can be seen as an insurance to reduce the risk of losing money because of bugs or user experience issues.
Is more thorough than what would normally be carried out by your team.
Is measurable in terms of savings, specifically:
Time freed up in the office,
Time saved fixing bugs post launch,
Cost of a full time dedicated resource versus cost of the service.

Will automated testing replace manual testing?

Tue, 22 Dec 2020 22:34:42 GMT

With the advancements in testing tools, automated testing has become very popular which begs the question “is manual testing still needed or beneficial?”

To answer this question let’s take a look at automated testing in terms of its advantages and disadvantages.

Benefits of automated testing include:

Testing and development can happen simultaneously.

Automated testing speeds up delivery times, as you don’t have to stop and wait for results.

Testing frameworks can be designed to fit your specific needs.

Just as there are benefits though there are also limitations to automated testing, namely:

There are lots of different tools available which can make it very difficult to choose one that is right for your needs.

Automated testing platforms are normally very expensive and can result in significant losses if you have not selected the correct suite of tools for your business.

Automated testing is not foolproof, i.e. bugs can still go through undetected.

Automated testing is only as good as the number of scenarios that have been scripted and is unable to report back on anything that falls outside of these “coded” parameters.

In addition to the above limitations there is a further limitation that, from my perspective, is the most important.

An automated test will provide your developers with useful information on each test scenario, but the feedback is one dimensional, i.e. value in, result = pass or fail. What it does not do is provide any feedback on how the user found the experience, or whether they were expecting the process or feature to work differently to what is presented to them. As we know, user behaviour is quite unpredictable at the best of time and while the scripted test delivers a pass result, from a user perspective this result could be a failure (Bug) as they were expecting the system to behave differently.

It is primarily for the reasons listed above that I believe end to end testing, specifically for websites are best performed by real people as opposed to automated test scripts.

Some people might argue that manual testing takes more time, requires a different skill set and an array of devices to test on. While this may be true, manual testing, especially if undertaken by someone who has not been exposed to the solution at any point during the lifecycle will provide you with invaluable feedback and insight to your solution. This insight exponentially increases the likelihood of your solution being better aligned to the intended user base. To further appreciate the benefits of outsourcing your testing, have a look at our blog post that covers this point in greater detail.

Automated testing is fundamentally important to the development landscape and I completely support the use of it in many cases. However, I do think that the one thing that automated testing cannot do well, is provide real feedback based on both the actual and expected results. This, as far as I am concerned is what makes manual testing a non-negotiable prerequisite for any product that is developed.