Why MENDELU throws 5.4 million crowns out the window

Today, let's take a look at the issue of tenders for digital projects. I would refer to this field as a royal consulting discipline. It is very difficult to make such a tender, which will really help the institution to find a great supplier.

And you may be surprised that this is often not so much because of the law as because of the inexperience of the contracting authorities.

For an excursion to this world, we will use tendering (VŘ) Mendel University in Brno, hereinafter referred to as MENDELU, announced on 16 February 2021 for ten new websites with a maximum volume of CZK 5.4 million.

In the article we will reveal why the tender procedure may not go the way MENDELU would like and what mistakes hundreds of other contracting entities are making. I believe that after reading the article you will never do them again and will take really experienced partners to help, such as 4E & Nehoupat (and no, I don't have a commission from it).

‍

How will it all turn out?

In the next chapters of the text, I will go through the MENDELU selection process from various aspects and try to explain why doesn't it turn out well for MENDELA, even if they eventually get those sites.

MENDELU has selected a tender for suppliers, but at the same time:

• does not share with the contractor why he is actually doing the project,
• it is impossible to say whether the project will be beneficial for the university (it can be said that new sites exist, but this is not the same thing),
• focus on graphics instead of content and technology
• the project is not anchored in the reality of qualitative and quantitative research,
• the assignment is very broad and requires work — this can take place during the project run, but it takes time that is not in the project today,
• the term is quite gallows even for one site of this size, let alone ten,
• the entire technical system will depend on one external programmer,
• suppliers are selected on the price and likability of references, which can not turn out well.

‍

And what's the worst? When MENDELU does it like this, no one will solve it (!) , since the sites have been launched, visually it will look better, so what's the problem? And in a few years, someone will think that there is a need to go to another redesign.

MENDELU today is expending energy, time and money in fundamentally the wrong direction and fundamentally wrong way.

Although I have a considerable amount of work done on digital projects of all sizes, I have written a book about it, lectured at all major events in our country — even then it is possible that I am wrong! So take a critical look at this article, just as I have viewed the IEC critically.

In all cases, I will talk about the risks of the project, which I perceive as uncaught and unaddressed. It is possible that MENDELU solves them in such a way that I do not see it from the tender documents and there is some internal coherence that escapes me. It can happen — but I will still take this opportunity to at least suggest to the other contracting authorities some aspects of the project that may negatively affect it.

Each of the listed areas can literally put the project on the shoulder blades.

👉 Poorly set goals

👉 Focus on graphics

👉 Pseudo-user research and assignment

👉 Sibenical term

👉 10 big bangs

👉 The programmer is driven by a car

👉 How to choose a supplier

I use italics for quotations from tender materials.

‍

1. Poorly set goals

Mendel University in Brno (hereinafter referred to as the “Client”) has decided to create new websites for the university, faculties and selected other university departments through the selected Provider in order to connect these ten sites into a new and common integrated solution.

The paragraph above is probably the closest to setting some project goals in the entire tender documentation. We decided to create 10 sites within one integrated solution. Sounds pretty cool actually, we launch the sites on deadline and the project goal has been met.

Digital projects have a catch. Just because I run a nicer website doesn't mean it will bring me anything. Plus, I pay for it, so it has a rather negative benefit -- until it makes me and gets paid.

The RIGHTER we do the WRONG thing, the WRONGER we become.

-- Dr. Russell Ackoff

‍

Why do universities have websites?

They want more applicants. Applicants who will ideally dream of going to this university/faculty from the third year of high school can choose a major and apply (ideally several times) and then enroll.

Universities don't have sites because of studentsbecause they have an information system and they go to the website as much as possible due to the opening hours of the study department. Other potential visitors are the general public, private sector partners and scientists. For simplicity, let's stay with the contenders.

Does MENDELA want to influence by redesigning the candidates' behaviour? That's actually not entirely obvious. It's possible that those 5.4 million just want to give for the technical migration of sites elsewhere, because they solve problems with the current solution provider threatening to shut them down, say. In that case, but it's unclear why it changes the graphics and content layout of some sites. Unfortunately, the specification of website visitors and their needs, concerns or motivations is completely lacking in the specification.

If the redesign of the ten MENDELU websites is to deliver anything, it is necessary to be clear in at least two areas.

1. What do we want to achieve?
2. Why should our approach to the project work?

‍

If MENDELU has set style goals make us 10 sites because we want themSo everyone will focus their efforts on making those sites exist and it's nice not to return the investment to the university in the form of more applications, for example.

If he wants MENDELA more applicants registeredas indirectly indicated by the demand for new graphics, content changes, and the close implementation deadline, the assignment doesn't follow at all why it should work. And this is also due to the fact that the contracting authority focuses on completely different things than those that could really affect the behaviour of the bidders.

‍

2. Focus on graphics

Each website is composed of three areas — content, technical solutions and graphics.

• Content is texts, photos, videos, infographics,...
• The technical solution is frontend, backend, hosting, domain and related processes.
• Graphics is the user interface (UI) and interpretation of the submitter's visual style.

‍

The entire selection process focuses on graphics that you will like to the sponsor. The evaluation criteria are built on the price and graphic references of the supplier.

• (the vendor will provide a functional url address and a screenshot (prinscreen) of the CMS administration interface with the WYSIWYG editor) — will be evaluated for clarity, usability and functionality;
• presentation of three realized web designs, the author of which is a member of the design team in the position of graphic designer according to Art. 6 para. 6.4, which the team member considers to be the top of his work and are thematically similar in focus as the subject of the public procurement (e.g. nature, ecology, university, etc.)

‍

In addition, the university will invite experts to help evaluate the supplied graphics to the smallest detail.

The Client points out that high emphasis will be placed on the quality of graphic designs and the unsatisfactory graphical appearance of the current pages is one of the reasons for the change. An independent expert from UMPRUM or a similar qualification may be invited to approve graphic designs by the client.

Paradoxically, there is less emphasis on the technical solution -- although it is described in much more detail. There is almost no emphasis on the content of the site itself. However, the importance of these three areas is completely opposite. If I want to influence the behavior of people out there, then the graphics have the least influence and the content the most. People don't go to the site because of the graphics, but because of the content.

It is possible that the university handles the content sideways within the framework of other deliveries or internal work. But in this case, it is curious that this fact is not reflected in the design of the structure of the site and the attached wireframes. Changes in the level of brand communication, web passes, information architecture, SEO or working with key objectives should be very well seen from them, but they are not.

The only thing we can learn about the content from the VR text — apart from wireframes and edits to part of the site structure — is that some of the content will automatically disappear.

The Client does not request the creation of new content for any part of the integrated website. The Client partially ensures the transfer of content from the original sites to the newly created web pages independently, while also requesting the machine migration performed by the provider through 20 conversion matrices in a total range of 1000 pages.

Focusing on graphics makes and does not make sense at the same time. The “competitors” of MENDEL in Brno are MUNI and BUT. Both have redesigned the visual style and with it the websites in recent years. MENDELU next to them today looks like a poor relative. But it's not the graphics of the site. It's a visual style, and a web graphic designer won't do much about it.

A web graphic designer really solves two things:

1. Responsive user interface (UI) elements. These are completely standardized today and all suppliers will supply them well enough and the like.
2. Interpretation of these elements into the visual style of the university.

‍

It can be done right or wrong. A top graphic artist will squeeze more out of the visual style than the submitter would have thought possible. But it does not change the fact that if a university wants a different graphic expression of its websites, it must primarily change its visual style NOT the graphics of the sites. This is just a result of the visuals.

If they hire a top web graphic designer, then of course the sites will be on a completely different level, and it can affect a certain part of the applicants. It has two hooks. A top graphic designer will refuse to work with the current assignment (see below), miss the deadline, and there will be no money in the budget for it. High-end graphics from a similar ecosystem of sites go for a million and it's just not in the budget. Not even half a million.

The bottom line is that for a visual style redesign, the experts from UMPRUM would be a real asset. Probably not for the redesign of the site. It is clear from the UMPRUM website that they have fields focused on traditional graphic design not digital design, and mixing traditional graphic designers into digital design is counterproductive in my experience. The site is not the poster and vice versa. Within universities today it is furthest in visual digital design UTB.

‍

3. Pseudo-User Research and Assignment

The Client has ready-made and legally approved wireframes of websites processed on the basis of user research, wireframes and their description are part of Appendix No. 2 — Functional specification of individual web presentations.

That's great! They did user research and then wireframes. The power of the contracting authorities does not do this to their detriment. We don't use wireframes much anymore, but they can be relevant and maybe even in this case they will be.

‍

“User Research”

So I clicked through more information and my heart stopped beating. Unfortunately, the user research submitted by MENDEL is not user research. It is the only and furthermore not yet elaborated questionnaire among students and staff. 1645 repliesthat don't help with a hint for the site redesign either. It is possible that the contracting entity has used other research methods, but does not share this fact in the assignment.

User research is a risk management tool. If we stick to the goal that the university wants more applicants, then the risk is twofold:

• Candidates are unable to use the site
• fail to influence them and thus increase the number of newly enrolled students.

‍

I expected Google Analytics data interpretation, keyword classification analysis, sense-making in-depth interviews or user testing. But a student-staff questionnaire? Why?

Current students are the least substantial target group for other college websites. The questionnaire is the least interesting method of user research when dealing with sites. I have no idea what the intention was. I have no idea what they got out of it. I couldn't get much.

‍

‍

“Wireframes”

Well, okay. They didn't do the research or hide it, so at least the wireframes. For your idea — university websites have thousands or tens of thousands of pages. Many of them are similar, but nevertheless there is some variability. There are twelve wireframes in the order (plus their mobile version). For the most part, it is least important pagesThis is the homepage of the site.

Sure, the site is called the main page, but half of the visitors never look at it, it's not an important part of the key passages through the site, and its content is actually of little interest to most visitors at all. A good homepage is a signpost. It's irrelevant in the context of the rest of the site.

You create and assign wireframes to eliminate the risk of a vendor delivering something that people don't understand or want to use. The main purpose of wireframes is to be testable. Two, whether they were tested or not, I have no idea. This is quite difficult to do with the main pages, and given that the content consists mainly of crossed out circles and does not even adhere to a clear typographical hierarchy, I do not suppose so.

The most important attribute of wireframes, mentioned several times in the EHR, is that they are approved. By the way, such fixation and sanctification by authority is probably the last thing large-scale digital projects need. The world is dynamic, it is necessary to change things according to what is happening outside and inside.

‍

“Information Architecture”

Let's look at the framework structure of sites. We can see from it what will be in the header as the top bar. And there are a few question marks, too. Does it mean that yet does not yet know how the applicant will look for his field of study?

Leaving aside the fact that the structure of the web and the information architecture are two completely different things, the proposed structure has a number of hooks:

• The structure is superficial -- when implemented, a large number of things will appear that will be different than the person in charge wants them (for example, in faculties).
• It does not reflect key passages through the site (one example for all are press releases, which virtually no one reads on other university sites).
• Nine links in a header is about five more than people need to make the site understandable and usable.
• I assume that the navigation proposed has not been quantitatively tested, which is standard for complex sites.

‍

In any case, I am glad that the authors at MENDELU like the section Life at MENDELU (taken from BUT) or the links Student and Employee in the right part of the header (taken from MUNI) . 😉

‍

Technical side of the assignment

The technical side of the sites is not my strong point, so I decided to ask Robert of Don't Swing. They do technical supervision of the implementation of the sites as well as the selection and I wondered what they thought about the assignment. There was a lot more to it than I quote in the following paragraphs. I left out the smiling parts.

‍

Robert:

The assignment is either recycled or plagiarized contracts from 2016 (unlike the MENDELU contract, a person other than the Contracting Party who participated in the preparation of the tender documentation is not mentioned here). Please compare Appendix 1 for both contracts.

Some paragraphs are also included in the contract of the Wallachian Museum in Nature in Rožnov pod Radhostem from 2019. All in all, statements appear in the assignment that are blindly copied between other submitters and it is uncertain whether the submitter even knows why he has them there (e.g. 2.4 Requirements for the design and functionality of the URL system).

The entire technical part of the assignment is then a poorly understood mix of functional and non-functional requirements full of conditionals, future time and inaccurate, sometimes smiley (“peak to 20,000 visits per day”) formulations. Very few technical indicators are defined by which the quality of the service provided can be judged. I found only 99.5% site availability (no definition as measured) and PageSpeed — unintelligible as one value for Web vitals, which make up multiple indicators, only Core Web Vitals are three indicators.

In addition, with a humorous objective (SLO):

The provider guarantees the speed of first loading of all types of web pages included in the wireframes measured by PageSpeed Insights at the level of overall Web Vitals rating for mobile devices 50-89 points out of 100 and for desktop 90-100 points out of 100. The Client explicitly warns that failure to comply with this qualitative criterion of loading the key pages of individual sites may be the reason for not accepting the work.

Does this mean that to enable the takeover, I prefer to slow down the site on mobile from 95 to 89 points?

Rather wild is then potentiality use the MENDELU infrastructure free of charge for the operation of the resulting solution. What bidder wouldn't use it when the hosting price is modeled over four years and factored into the total price, which has a weighting of 30%?

Further support requests are then either laughable (critical defect = defect that threatens the good name of the customer, existence of very critical defects...) or downright ridiculous: “The Provider shall provide the Client with a copy of the Website, including the program key, code, database and all data so that the Orderer can restore (operate) the Website even without the active assistance of the Provider. “ Of course, there is nothing ridiculous about the goal behind this requirement, but in this way it is not really possible to fulfill that goal.

In general, technical requirements are very inaccurately described, chaotically arranged and poorly controllable. Moreover, there are desperately few for a project of this size.

Thank you Robert for the comment and let's move on.

‍

4. Breaking term

We don't have assignment, we don't have intent, we focus on inappropriate areas, but we have a deadline. And that term is indeed gallows. The contractor has to create 7 new sites in 4 months (university and faculty), followed by 3 more sites within 2 months. I suppose he will fill the gaps in the assignment with his own imagination, so let's not solve them gracefully and let's go to the deadline.

It boggled my head for a while as to why the term is so cringeworthy. The only reason I've come to is that the university wants to be ready for the season with a new site. The season starts in November and the website is a key touchpoint for communicating with applicants thanks to the pandemic. All 10 sites must be in full force and ready to lure and entice thousands of applicants.

I imagine at some point there was a dialogue between HIPP (Highest Paid Person) and the person who got the sites in charge. Let's call her a project manager.

HIPP: “We need bidders to see our new site next season.”

Project Manager: <Mlčí>

HIPP: “I'm sure you can handle it. “

Project manager: “So we'll put it in the contract and everything will be fine. And we will ensure this with a penalty of 5,000 CZK for the starting day of delay.” <pro mě cajk, kdyžtak to hodíme na dodavatele>

Let's break it down:

• Even if the contractor prosecuted everything, the university simply does not have a chance to respond and provide synergies so quickly. There will be no more endorsements, moreover, across the faculties, for a week. And without the cooperation of the university, new sites cannot be launched.
• There's no critical path in the assignment that says the term is real. I don't believe this risk is managed in any way, other than outsourcing to a contractor who signs it, the deadline for some reason on the university's side doesn't work out and nothing happens.
• Making a good design system is not a matter of the week. For example, at MUNI we debugged and improved the graphics with the supplier of the unit of months.

‍

If, on the part of the university, I was dealing with the season and attracting applicants, the last thing I do is redesign sites. For example, a BUT does a great creative campaign every year that attracts an order of magnitude more attention than their gradually redesigned sites. A different country, a different kind. MENDELU instead awaits a big web bang.

‍

5. Ten Big Bangs

The Big Bang means that a large number of changes occur at the same time. For example, we will change the technical solution, the graphics of the site, and we will also intervene in the content. And then at a critical moment (like the beginning of the season) we start everything and hope that there will not be a single mistake in our plans.

The bad news is that there will be bugs and changing sites will have unexpected consequences. And MENDELU expects this tenfold. The complexity is at the level of logistics of vaccination against COVID in the South Moravian Region. Risks are everywhere.

• The web can fail technically — even if we don't know it at first glance
• Repair of some parts can be considered as multitasking due to lax assignment, so it will take place next year (the contractor will have his hands full with the fulfillment of the contract, and there will be time for multitasking later)
• Candidates or other targets do not have to navigate it
• It may not be convincing enough for them
• It can and will break down on different devices
• There will not be a lot of minor details that will make people unable to fulfill their agendas that the contracting authority today has no idea about
• Content editing won't be finished—that is to say, we'll be removing content from the existing site, but it's never that simple, especially when graphic creations are expected and our now-already-long headlines flow out of them
• Some faculties will not want a new website, they will passively oppose it and start making their own websites
• ...

‍

And all of this will come AFTER the new sites are inaugurated, so there will be no turning back. The contractor will remain as a normal business and will have the adjustments paid for from a pre-approved budget.

Only one big bang poses a risk, let alone ten. As we begin planning a redesign of a larger site today, we are not planning it for this Christmas, but for the next one. If it is needed earlier, we choose a different path. MUNI has about 800 sites, BUT about 150. Neither university launched them at the same time. One was triggered. Unexpected bugs have been debugged. The team took a breather and went on to the next one. The parallel solution of multiple sites at the same time occurred only after some time.

What is the situation with MENDELU's implementation team?

‍

6th. The programmer gets hit by a car

The implementation team will consist of:

• Projektový manager (1 person);
• Grafický designér (1 person):
• Backend programmer (1 person);
• Frontend programmer (1 person)

‍

Who will handle UX or web analytics is unclear. But let's take a different risk. All MENDEL sites will run on a system that will be managed by a single programmer. Sure, there may be more, but they DON'T HAVE TO. What if she dumps him for a month of COVID? Where's the advantage over the state of having my own programmer as an employee? The required fungibility is zero.

In the contract, it is possible to find a good effort to fix key team members, but with a rather problematic solution:

8.2.3 The Client reserves the right to refuse significant changes in the composition of the Implementation Team at the time of performance of the Contract.

What happens if a programmer quits or gets hit by a car, hell knows.

The Provider will organize all development in a suitable versioning tool (such as Git) and will transfer live access to this versioning tool to the client's authorized persons. In addition, the Provider shall keep documentation of the work in an appropriate form in such a way that it is understood not only by him, but also by authorized persons of the Client and, if applicable, by a third party authorized by the Client.

The Provider undertakes, upon request, to transfer the further management and development of the Sites to the Client or to a third party authorized by the Client.

On the other hand, the contracting authority strives for the transferability of the work to a third party. In practice, the hope of passing on a project is small, as with all other IT projects. Despite all the efforts, it may turn out that the documentation is outdated and the code should have been checked by someone, but in fact it does not. And breaking into foreign code is always painful and quite expensive. For example, for a certain e-shop, the take-over price was twice that if someone did it from the beginning again.

Probably the safest option is to solve the system by yourself through an internal team with sufficient fungibility. At least that's how both MUNI and BUT address it.

‍

7. How to choose a supplier

I keep the biggest goal in my own goal at the end. The method of choosing a supplier. Let's start with the evaluation criteria. There are three in total. The first two are the price — for creating 10 sites and developing them. Interesting is the third criterion, which accounts for 40% of the weight. It is named Quality of the proposed solution.

A more accurate title might sound Likeness of reference projects. In fact, the evaluation committee's feeling from the previous graphics references and the perceived visual quality of the administration will be evaluated.

• a functional sample version of the CMS administration interface, including a functional WYSIWYG editor, which the vendor proposes to use (the vendor provides a functional url address and a screenshot (prinscreen) of the CMS administration interface with the WYSIWYG editor) — will be evaluated for clarity, usability and functionality;
• presentation of three realized web designs, the author of which is a member of the design team in the position of graphic designer according to Art. 6 para. 6.4, which the team member considers to be the top of his work and are thematically similar in focus as the subject of the public procurement (e.g. nature, ecology, university, etc.) and an understandable presentation of the author's directions and intentions during their work creation — the implemented designs will be evaluated in terms of clarity, usability and formal order.

‍

The documentation comprehensively describes the clarity, usability, functionality and formal order. They will be evaluated on a three-step scale.

• 0 points = unintelligible/complex to use/missing important functionalities/formally disordered;
• 1 point = intelligible/usable/functional/formally arranged;
• 2 points = extremely intelligible/extremely easy to use/extremely functional/extremely formally arranged.

‍

Do you see that too? Genuine feelings! Any supplier can challenge this and the VOR can drag on for all the time that MENDELU has prepared for implementation. I've been doing websites since 2001 and really I have no idea how an intelligible proposal differs from an extremely understandable one and I wouldn't dare distinguish it. Likewise -- sites have different target audiences, and the review board doesn't have a patent on sanity.

Again, I decided to ask RobertWhat does he think about it:

The evaluation of references is at least on the edge of the law — although the EEZ does not explicitly prohibit the evaluation of the qualifications or experience of the supplier compared to the previous legislation, the professional public is rather inclined to the immutability of this prohibition also in the context of the OCTs. The qualifications of the supplier, in particular the quality of the reference contracts, do not, according to those views, relate to the qualitative aspects linked to the subject matter of the public procurement. For fine artists, see Dvořák, D., Machurek, T., Novotný, P., Šebesta, M. et al. Public Procurement Act. Comment. 1st edition. Prague, 2017

It is good that contracting authorities are gradually trying to involve quality in the evaluation, and not just to evaluate purely on price. Unfortunately, this is a demonstration of how it can't be done I can't imagine how the sponsor will credibly justify his decision. The results of such competitions will rather serve to make other contracting entities afraid to evaluate quality. Finally, the contracting authorities make excuses that the Public Procurement Act does not allow this and return to a pure evaluation of the tender prices, which have nothing, especially in our industry, to do with the evaluation of the true economic benefits of the offers. While such evaluation is also possible under the current regulation of the law, contracts using the BVA (Best Value Approach) e.g. in implementation M4E by 4E Consulting can also be found at Brno universities (MUNI, BUT).

The whole selection process is then set up quite unfavourably (deadlines, criteria) and in the spirit of “the contract must be a deterrent”, which has persisted for some contracting authorities since ancient times, when companies fought for public IT contracts.

Similarly, other deterrent provisions such as the setting of penalties, guarantees,...

The assessment of the contract could (as well as the assessment of the Procurement Documentation — evaluation criteria, etc.) be published for the next article. I've already dissected myself enough. So let's bring out something at least a little funny:

10.9. As a password for verifying the legitimacy of reporting a request pursuant to Article 4 of Annex 1, the Contracting Parties agree... [TO BE ADDED BY SUPPLIER].

It remains only to hope that someone will not forget to darken the password before publishing it in the register of contracts. 😃

‍

What to take away?

Doing a good assignment is no fun. There are several ways out of this. I would definitely start by adding MENDELU:

• What is the goal of the site?
• Who is this site for? Who is the most important target?
• What are the key scenarios and conversion actions of each target?
• How does it differ from site to site?
• Why should our ideas work? (except that we copied it from elsewhere and it probably works for them...)
• ...

‍

And subsequently changes the evaluation methodology, the contract and sharpens the technical requirements.

With the assignment and contract currently set, they can't expect to get a good supplier. even if everything goes according to plan, the risk is so much that it will cause university representatives a real headache. And that for the next few years.

‍

FAQ

‍

Why do you write stuff like that?

I don't want public procurement officers throwing money out the window. I don't want that with companies either. I understand that the popularity contest is probably not going to win me the article. It's long, heavy, and for most people actually kind of boring. I am pleasantly surprised that you are reading this paragraph.

‍

Do you think the assignment was done by morons?

Not by chance. I assume there are people working at the university who are specialists in something other than web design and digital project tenders. It really seems simple until you get into it.

‍

Why do assignments for digital projects turn out like this?

Institutions have procurement departments and lawyers, but the latter are used to competing toilet paper on price rather than big digital projects on quality. So on the one hand we have a progressive employee who would like to compete on quality. And he is caught between the millstones of established procedures and the impossibility of choosing experienced partners for the HR, because we have lawyers, so what would an internal audit say?

The result is a heap of objections, complaints against the ÚOHS, cancelled tenders, delays and, at worst, fines. And unfortunately, also the burned-out employees who were just trying to do their job well.

‍

Is this assignment completely wrong?

Ah yes. In addition to the fact that I do not consider the assignment to be good, the tender documentation itself and the contract do not minimize the risks associated with the selection of the supplier and do not help to make the chosen partner a good partner.

‍

Is the $5.4M a relevant price?

In the estimation, the amount does not depend on the actual state of affairs, but is set so that the order is below the limit. If I'd take it a lot over the surface, they want 10 sites, hosting + maintenance for 48 months, and 2400 extra hours of work. I'll keep it low:

• 2400 h of work is say 2.4 M, more like more.
• hosting will be free from MENDELU
• maintenance and support for 48 months is say 30,000 per month, so 1,440,000 CZK
• There are about 1.6M left for sites, that is 160,000 CZK per site

That's actually quite a bit. My understanding is that the submitter doesn't want 10 sites, but one big system to build the sites on. But even so -- those sites are not small, and a large number of question marks remain to be resolved.

‍

Where do we start?

Graduate from our free email web project management training. You can find it at below.