Google Shopping

Google Shopping Tips & Tricks Part 3: Bidding & Strategy

After you’ve optimized your shopping feed and configured your campaigns, it’s time to move on to bidding and strategy. This process can be easy or difficult, depending on the type of campaigns you’re running and the internal capabilities of your team. Remember that none of this is finite. If performance isn’t good or you’re struggling to utilize budget wisely, you can always pause everything and get help from an agency or try something that requires less management.

Smart Shopping

If you decide to try running a Smart Shopping campaign, you’re in luck! Almost every single facet of this campaign is automated by Google. There are only four areas where you have control:

  • Daily budget
  • Target ROAS
  • Products in the feed
  • Creative assets

 

It’s hard to advise on budget, because all merchants are unique and have different goals. Typically, I recommend allocating 30-40% of the overall paid search budget to Google Shopping as a starting point. Some merchants may spend up to 60% of their budget on shopping initiatives; it all depends on your goals and what works well during the testing phase.

Once you’ve set a budget, Google will automatically maximize your bids in order to spend the full amount. After running your campaigns for 3-4 weeks, assess your average ROAS. Let’s say that it’s hovering right around 3.6x after the fourth week. At this point, you should go back to your bid settings and configure a target ROAS goal. I typically set this slightly higher than my average return. For a 3.6x average ROAS, I would set my initial goal to 4x. After doing this, you might notice that Google no longer spends your full daily budget. If spending the entire daily budget is a priority, simply lower your target ROAS goal.

As previously mentioned in part II of this series, it is possible to run multiple Smart Shopping campaigns that contain different products, or a mix of Smart and Traditional campaigns. You don’t have to run this type of campaign for your entire product catalog. To figure out what works best, test as much as possible and be sure to document your results each week.

The final area where some control exists is with creative assets. Unlike Traditional Google Shopping, Smart Shopping shows your products in a variety of placements and allows a wider range of asset types (static images and logos, along with product images). Adding and testing different assets is always recommended.

If this type of campaign sounds a bit too basic, Traditional Google Shopping might be more your speed. The primary reason we see clients shy away from Smart Shopping is concern over ad placements. It is very possible that ads could show on off-brand websites, and since Google doesn’t provide good information about any of this, retailers would never know for sure. If brand integrity is a major concern, stick with Traditional Shopping.

Traditional Shopping

Traditional shopping is far more complex and requires significantly more planning and management. Unless you have an in-house expert or plan to work with a very small budget, SD recommends hiring an agency to take on this initiative as it can be very time-consuming.

Managing bidding and strategy for Traditional Shopping could easily be its own three-part blog series, but here is a brief overview of the most important insights I’ve learned over the years:

  • Organize campaigns by high, medium, and low priority settings. These campaigns should use different negative keyword lists and match with different types of queries:
    • High – generic queries that don’t contain your brand name.
    • Medium – queries that contain your brand name.
    • Low – queries that contain specific product names or SKU searches.
  • Utilize a mix of automated and rule-based bidding. After identifying products that don’t have a great return, SD recommends separating them and testing each automated strategy to figure out which works best – maximize clicks, enhanced CPC, and target ROAS.
  • Adjust bids based on audience type and take advantage of lookalike groups. If a user is more likely to buy something, bid up and increase visibility in order to win their business.
  • Test geo-targeted campaigns. If certain cities perform better for you, break them out into a separate campaign with a more aggressive bidding strategy.
  • Test new features. For example, showcase ads contain multiple products and often have a higher CTR. Try running these to see if they work for your business.

 

  • Take advantage of promotions, customer and store reviews, and competitive pricing/shipping. These features all increase CTR and drive customers to convert.

 

With Traditional Shopping, the best advice I can give you is to test everything, carefully document results, and slowly optimize. Success takes time and there will certainly be some pitfalls along the way. What works for another retailer may not work for you and you’ll never find out unless you try.

As always, don’t hesitate to contact us if you need assistance or have questions. Good luck!

Written by: Lindsay Pugh, Senior Digital Strategist

The Third-Generation of Paid Search is Here  

Like many of you, I started my career in the dotcom era, when ecommerce was first becoming “a thing.” Retail sites were launching fast and furiously, and Wall Street couldn’t invest in them quickly enough. E-tailers had it easy back then. Customer acquisition costs were low, and paid search was like an ATM that kept dispensing money without ever requiring a deposit. Every one dollar spent on paid search delivered seven or ten dollars in sales. Those were the days!

Back then paid search meant buying text ads in Google, which, by default, was the place that all users began their purchasing journeys. Soon enough, all e-tailers learned that paid search was essentially a gravy train and so they jumped in, which of course, drove up customer acquisition costs and lowered ROI. It was the end of an era.

Then in 2007 Google acquired DoubleClick, along with a bunch of search and display networks, and ushered in the second generation of paid search, which centered on banner ads placed on relevant sites. The theory is that if I’m reading an article on Italian cooking, I would be open to seeing an ad for olive oil. These ads, purchased on a CPM basis, didn’t deliver great ROI, but in aggregate made money for the advertiser. Some retailers did extraordinarily well.

Second-generation paid search led to some important innovations in ad technology. First and foremost is retargeting, aka lead-back campaigns, which has been the gold standard in ecommerce for the past eight or so years. Here’s how it works: users visit your site, perhaps look at items, and you site drops cookies into their browsers. Once they leave your site, they see ads for your brand or the items they viewed wherever they go next, whether that’s the New York Times or Instagram.

It’s not an exaggeration to say that digitally native and D2C brands owe their success to this tactic. By spending a relatively small amount of money, tiny startups with no name recognition, — Hims and Hers, for example — can elevate their brands by simply being ubiquitous. Let that sink in: for consumers who visit a site once, Hims or Hers takes on the same prominence that brands like P&G spent decades and billions of dollars building. This marked the end of the second-generation of paid search, in my mind.

We’re now deep into the third-generation of paid search, which I think is rather unfortunate. Consumers no longer begin their journeys from a single place, but from many: Google, Bing, YouTube, Facebook, Instagram, Pinterest, Amazon, Alibabi, Walmart, among others. Now, in addition to the text and display ads, retailers need video for the highly visual channels, such YouTube pre-roll ads and Instagram interstitial stories. And, like banner ads, video placements come in all different sizes and placements; full screen for YouTube pre-rolls, landscape or square for Facebook feeds, and squished down to fit in a banner or side ad on a news site. Campaigns today need hundreds of creatives in multiple formats and sizes.

The real challenge to third generation paid search is that it requires a multitude of strategies, accommodations and budgets. For instance, Amazon has its own paid search, and Google now has Google Shopping, which means retailers must ensure that both Amazon and Google understand their product catalogs so that those platforms can serve up the right products. In other words, retailers need to format their product data in ways Google, Bing, Amazon, Facebook, etc. can understand it.

To be a marketer nowadays is incredibly difficult. Customer acquisition costs reach higher into the sky while ROI is on the decline. One can’t help but wonder if there is a fourth generation of paid search, and if so, will it lead to more or less fragmentation. If more, what will that mean for marketers?

Clearly, retailers, especially SMB companies can’t be everywhere, which means they need to make strategic decisions. Which channels make sense for your brand? Where on the web does your target audience congregate? Which ad formats work best in those channels.

These are questions Something Digital spends a lot of time talking about with clients, which is one of the many reasons why we developed BULLSEYE, an ecommerce business maturity quiz. We can help you assess the maturity of your business, which in turn helps you make strategic decisions around where you should invest.

At the moment a lot of our customers having good success on Google Shopping because its reach extends beyond Google. For instance, Google Shopping results appear in YouTube. For many retailers, that’s a perfectly fine near-term strategy.

If you have questions about paid search or need help with it, contact us.

Written by: Phillip Jackson, Ecommerce Evangelist

5 Things That Scare Customers Away from Ecommerce Sites

New customers are the lifeblood of online retail, which is why ecommerce teams invest significant resources in attracting new users to their brands. To wit: the IAB reports that of the $107 billion spent in digital advertising in 2018, 45% went to paid search. That’s just under $50 billion. So when you succeed in attracting new users, the last thing you want is to have them leave, never to return. What scares them away?

This is a topic that Something Digital investigates on a continuous basis. While the exact set of reasons differ by industry segment and customer type, we can say that on the whole, customers often abandon a site for one of the following five reasons.

#1: No Sustainability Messages

Today’s modern brands — Everlane, Allbirds and Casper — have an outsized influence on consumers. These brands have tapped into the consumer’s desire for ethically produced products, and feature their commitments prominently on their sites.

It’s a sound strategy. According to Nielsen, more than 80% of consumers say it is “extremely” or “very” important for companies to implement programs that improve the environment. And that concern for the environment has a direct impact on their behavior. How do we know this?

Recently, Something Digital systematically looked at our customers’ websites and discovered that the About Us page had the highest bounce rates. This is the place where modern, environmentally conscious consumers go to when they want to see how green a brand is. They leave the site if all they see is the founder’s story.

Sustainability may not be your top-line marketing message or even something your brand spends a lot of time thinking about, but the world is changing around you. Direct-to-consumer brands are doing for sustainability what Amazon did for package delivery, namely setting the bar for consumer expectations.

#2 Not Sure What Happens After They Give You Their Money

Your business might not be able to match Amazon’s next day or two-day delivery, but that’s okay. What’s not okay is leaving consumers in the dark about what to expect. Will their product arrive in two weeks? Will they receive a notice when their product has shipped? The truth is, most consumers don’t mind waiting, especially for considered purchases (and especially if they’re aware of the brutal impact Amazon’s next-day delivery service has on the ware employees who must fulfill those orders). They just don’t like the unknowns.

The site Sweetwater excels at setting expectations. The product pages for all of their products provide arrival information, which goes a long way in building trust. Shoppers know when they complete a purchase exactly when they can expect the have the item in their hands.

#3: Decision Fatigue

Choice is great, but too much of it can easily overwhelm customers, causing them to leave your site. I’m a fan (and loyalty member) of Nordstrom, but I admit that shopping for shoes on their site is a miserable experience. When I sort by my shoe size and then click on a subcategory, I reasonably expect the site to present me with the options that are available in my size, but that’s not the case at all. Instead, I’m dropped in the middle of hundreds of thousands of items, and it’s up to me to which come in my size. So even though I’m a loyalty member and it was the Nordstrom Anniversary sale, I left the site.

Guided selling is an ecommerce tactic that Something Digital can’t endorse strongly enough.  Guided selling comes in multiple flavors: quiz-based personality testing, questionnaires, or even chatbots that walk consumers through a decisioning tree. Regardless of the format, the goal is always the same: help the customer find a product that is right for him or her.

Check out our new Guided Selling eBook, produced in conjunction with Zoovu.

#4: Comparison Shopping is Too Hard

Cross selling is a tried and true tactic to stem abandonment. If one product isn’t quite right, or above the consumer’s budget, the retailer can still win that customer by presenting a similar one, perhaps with a different feature set or price point.

It’s a great strategy, but only if it’s easy for consumers to compare them easily. All too often, that’s not the case. Typically, when consumers look at one product and then another, the site will make it easy for them to find the original item again, but they can’t look at them side by side.

Sales will go up if consumers can see multiple options all at once, as Sweetwater does with its guitars. This should be the norm:

#5: Lack of accessibility

At some point in our lives we all experience impairments that prevent us from using a website or mobile app. We may have misplaced or glasses, or are trying to order a pizza from a mobile app but can’t see the screen in the bright sunlight. When that happens, we click away for a more friendly site. Some consumers will even sue the site.

In 2018, nearly 2,300 ADA lawsuits against businesses were filed in federal courts, claiming their websites weren’t ADA compliant. And it looks like Domino’s Pizza is going to take its challenge all the way to the Supreme Court, given that this past January, the Ninth Circuit Court of Appeals confirmed that the ADA applies to the chain’s website and mobile app.

We at Something Digital have long held that the benefits of creating an accessible site are manifold. By offering users the ability to access larger text sizes or by paying attention to contrast, your site can be used by a greater swath of people, including those who are temporarily impaired.

A best practice is to ensure our site visitors can always find what they’re looking for, and that means adhering to WCAG standards.

Get in Touch

Do you know why visitors are leaving your site? Do you have a strategy for reversing that trend? If not, send us a message.

 

Written by: Phillip Jackson, Ecommerce Evangelist

Compete Amazon

Budgets are for CFOs, not Paid Search Campaigns

Last year I heard a business partner of ours, David Deppner, say something both startling and incredibly insightful at a conference. David is cofounder of Psyberware, an ad management solution for Magento merchants, and his company (among other things) helps merchants drive efficiency in their paid search campaigns. So when he said, “budgets are for CFOs, not paid search campaigns,” I sat up.

Some context: the discussion at the moment was focused on customer acquisition, and David noted that every marketer is eager to achieve specific customer acquisition costs (CAC) and ROI, and to that end put limits (i.e. a budget) on paid search. This doesn’t really make sense.

Look at it this way: If you knew that every dollar you put into the bank grew into three dollars in six month’s time, wouldn’t you keep depositing your money there? This is paid search. It’s the investment that keeps on growing.

More than that, paid search is all about growing your top funnel. Unlike remarketing or retargeting campaigns, paid search is focused solely on customers to your brand and ecommerce store. Spend a dollar acquiring them and you’ll get your dollar back in the first sale, and additional dollars when they return.

True, many retailers, particularly direct-to-consumer brands, face high customer acquisition costs. But they also have high retention and loyalty rates, which directly translate into repeat sales, which justifies the higher margin they’ll need to pay to capture the customer initially.

I’ll give you a concrete example. Let’s say you’re a shoe brand and one of your evergreen products is a $138 pair of leather sandals, and your average order value is somewhere between $110 and $120. Spending $40 to $50 to acquire that customer is quite reasonable given that you will probably make 25% margin on those sandals. When all is said and done, you’ll get a respectable $35 from the initial transaction, but that’s just the first benefit.  Customers you sell to are 60% to 70% more likely to purchase from you again, and many are likely to recommend your brand to friends and family.

Unfortunately, far too many retailers view paid search as an expense to be controlled and capped, and not as an investment in building the top of their sales funnel, and ultimately the future success of their businesses. Assuming you’ve built a smart strategy and have partnered with the right companies, in theory you should have an unlimited pool of money at your disposal for your paid search campaign. By this I mean, for every one dollar you put into paid search, you stand to add two to three dollars in future sales revenue.

So why do retailers have budgets around paid search spend? Now a cynic may say this is rather self-serving, as Something Digital is a digital marketing company and probably doesn’t want to work within a budget, but that’s not the case at all. We certainly wouldn’t recommend that any company spend more on paid search than it has in the bank. What we are saying, however, is that most budgets allocated are on an accrual basis, in the same way that  PTO is accrued by a full time employee. But that’s not how marketing spend should align.

An intelligent approach is to plan how you are likely to spend your marketing budget over the course of the entire year and to know upfront when higher paid search costs are likely to deliver higher dividends. Case in point: gifting holidays provide the opportunity for retailers to gain up to three new customers from a single customer acquisition. This means if your AOV is $110 and you spend $60 to $70 in paid search, your ROI is lower the first sale, but much higher when you consider $220 from two new customers you didn’t need to pay for. Abiding by strict CAC (customer acquisition cost) and ROI goals during peak gifting periods means you miss out on this upside.

To CFOs, budgets are tools to track performance and assess whether or not the company is investing in the right technologies, processes and market sectors. Without it, CFOs have no real way of knowing if they can meet their financial commitments and responsibilities. That’s not the marketers role.

The marketer’s job is to strike while the iron is hot. When you impose limits on customer acquisition, you needlessly limit the maximum amount of performance you can drive in a calendar year. This isn’t the best way to approach marketing. If Allbirds had taken that approach it wouldn’t be a unicorn (valued at $1 billion) today.

That’s why every marketing budget should offer flexibility to adapt with the demands of the marketplace. It is, after all, the future of your company, and you shouldn’t put a cap on its long-term success.

 

Written by: Phillip Jackson, Ecommerce Evangelist

Guided Selling Ebook

Since the dawn of ecommerce, retailers left it to the consumer to figure out which products were right for them. They provided a search function, a product image and details (often taken verbatim from the manufacturer) and called it a day. It was consumer’s job to slice and dice the catalog and to arrive at a decision. True, some offered live chat assistants to guide consumers, but mostly they followed pre-written scripts, and lacked the kind of detailed knowledge needed to help visitors select products.

Guided selling seeks to address the “decision fatigue” consumers feel when presented with too many product choices. Guided selling takes multiple forms: quiz-based personality testing, questionnaires, chatbots that walk consumers through a decisioning tree. Regardless of the format, the goal is always the same, which is to help the customer find a product that is right for him or her.

Want to learn more about Guided Selling? Drop your email in the box below for our free Guided Selling Ebook.

Written by: Something Digital and Zoovu

 

 

Enter your email below and receive a copy of this eBook in your inbox.

Your eBook is on the way!

Bullseye Series Part 3: Partners Pave the Way

Earlier this summer we announced Bullseye, a quick and easy assessment tool we designed to help B2C and B2B companies select the right partners and services to take your ecommerce business to the next level. This three-part blog post series looks at Bullseye from different angles:

  • In Part 1 we discussed how we use Bullseye to measure the maturity of a business (it has nothing to do with revenue or business acumen and has everything to do with the size and skillset of your ecommerce team).
  • In Part 2 we talked about how to right-size the tactics and strategies you deploy to grow your business; how to define and measure success; and how we use Bullseye to determine both.

 

In this blog we focus on how we use Bullseye to right-size your partnerships. This is critical, because through years of experience we’ve learned that ecommerce businesses can succeed or fail based on the tools and the partnerships they form to engage their customers.

Right-Sizing Your eCommerce Platform

When some ecommerce managers set out to build their infrastructure, their first impulse is to opt for the platform with the most bells and whistles. This approach, their thinking goes, allows for maximum flexibility in the future. They may not need all the features and functionality now, but who knows what will happen five years from now?

In reality, it may be like buying a Ferrari 458 just to go down the driveway to get the mail. It’s way more power than you will ever need for the task at hand and a huge waste of money.

Though it may sound counterintuitive, features and functionality don’t necessarily equate to business success. Just because you can do something it doesn’t mean you ever will or even should. This is why Bullseye and right-sizing is so important. You can choose (and pay for) an ecommerce platform based on its impressive list of capabilities, just take care not overestimate the ability of those features to drive value for the company. Case in point: among Magento’s hundreds of features, is a wedding registry, but how many companies will actually use it? Certainly no B2B company will ever need such a feature, and gift registries don’t make sense for many B2C companies.

We at Something Digital are huge fans of Magento’s ecommerce platform. We’ve been an enterprise-level partner for 10 years, and have built hundreds of sites on it. Personally, I’ve worked with Magento since 2007. Twelve years ago the company offered features and functionality for the B2C market that are on par with what competitive platforms of today. The company was ahead of its time.

While we’re fans of Magento, we don’t recommend it for high-growth but smaller staffed B2C merchants. Magento requires more internal staff, along with an understanding of generally accepted accounting principles, order management, order lifecycle and so on. Nor does it come with a preconceived notion of how your business should manage orders. If your ecommerce team is small, and doesn’t have the skillset to map out an order management system, we’d recommend Shopify over Magento.

The same is true for any software used to support your ecommerce business. Does it make sense to spend $3,000 a month for an email platform if all your team has time to do is send out newsletters? You don’t need the Ferraris of software if you don’t have the capacity to utilize all of its capabilities. You’re better off spending your money elsewhere in the business.

When evaluating any software choice — whether that’s an ecommerce platform, email service provider, CRM or ERP system — you need to ask:

  • Does this software meet my needs today?
  • Does my company have sufficient maturity to utilize the totality of this software
  • Do we have the vision to utilize this software to grow the business within a reasonable time frame?

 

These questions can be difficult to answer, which is where Bullseye comes in. We’ll assess the maturity level of your business and recommend partners who will be a good fit for your company.

Role of Bullseye in Recommending Software

There are a lot of things to consider when assessing the maturity level of an ecommerce business, but that assessment is absolutely essential in order for Something Digital to recommend the right partners to engage. We can’t say this enough: ecommerce maturity isn’t about gross revenue, it’s about resource levels and skill sets of team members who will be tasked with growing the business. Choose the wrong partners and your business can suffer; select ones that compliment your team and infrastructure and your business can prosper.

Interested in obtaining a free Bullseye assessment? Contact us now.

Written by: Phillip Jackson, Ecommerce Evangelist

Bullseye Series Part 2: Growth, Goals and KPIs

In a previous post we told you about, Bullseye, our new tool to measure your business’s maturity so that Something Digital can help you select the right partners and services to grow your business. But that’s just the start. An important goal of Bullseye is to identify the best areas to target for growth, and to put a rational strategy in place to succeed.

One of the things we say a lot these days is that the outcome of any strategy should be set of specific tactics that work towards the goals set. It’s why Something Digital is hyper focused on tactics. At the same time, we need to measure success. There’s no point in putting a strategy for growth in place without a means of measuring whether or not it’s effective.

This is why Bullseye is so strategic: it helps to ensure that Something Digital recommends tactics that will advance your growth strategy, and are appropriate to your company’s infrastructure, available resources and skillset.

Let’s say your top goal is to increase your conversion rate. A worthy goal, to be sure, but one that for more mature companies must be broken down into separate KPIs. Here’s why: every website has new, intermittent and loyal customers, all of whom convert at different rates and timeframes. Focusing on a single metric — your overall conversion rate presented in Google Analytics — won’t allow you to put tactics into place that are geared towards each user type. To really grow your revenue, your strategy should include granular tactics.

Our approach is persona based. We look at your site metrics to identify personas of customers for every stage of your sales funnel (that makes sense for your level of maturity), and then create specific, measurable tactics for each. Such an approach requires multiple tactics and KPIs. In fact the more mature the company, the more narrow the KPIs to measure. Growth will come from understanding the entirety of the customer journey, as well as tracking and optimizing all of the touchpoints prospects engage with prior to making a purchasing decision.

Take first-time visitors. A KPI we might want to track is the length of time it takes for these visitors to purchase; a goal may be to narrow that window; and a tactic can be to prompt a smaller conversion, such as encouraging the visitor to opt in to the site’s newsletter. Another KPI we may want to track is increasing the rate of return visits by new leads; a goal may be to spend remarketing budgets as efficiently as possible; and a tactic can be a lead-back campaign that’s executed across social media.

On the opposite end, a level-one Bullseye company will require a different set of KPIs, goals and tactics. For instance, level-one companies may not have enabled Google Analytics advanced ecommerce metrics. If that’s the case, our first priority is to ensure they have the means to define and measure success. An ecommerce team can work hard to increase conversions across the site, but if they can’t measure it, they can’t claim it.

Next, we’ll need to identify where and how the company can drive the best returns on the investments they make. Often, the best ROI stems from focusing on visitors with the lowest conversion rates. Let’s say that mobile visitors convert far less frequently than desktop visitors, we would focus on optimizing the mobile experience to get more conversions right away.

Some tactics can apply to all levels of maturity, although with different implementations. For instance, if the goal is to increase conversions, a tactic we may suggest is adoption of a conversion rate optimization tool, such as HiConversion (an important partner of Something Digital, and a participant in the Mobile Optimization Initiative). HiConversion offers a powerful A/B testing tool that benefits businesses of all maturity levels, but how it’s used will vary.

At the end of the day, Bullseye is designed to help Something Digital:

  • Identify, track and optimize appropriate KPIs
  • Right size tactics to deliver on the strategy
  • Ensure the strategy takes your business to the next level

 

Next up: How Bullseye Paves the Way for Partners.

Written by: Phillip Jackson, Ecommerce Evangelist

 

Bullseye Series Part 1: Measure Your Maturity

Something Digital has a new tool that’s super helpful to ecommerce businesses who are outgrowing their websites, considering building a wholly new one, or who generally want a health check to ensure they’re sites are working optimally. That tool is called Bullseye.

Bullseye is a quick and easy assessment tool we designed to help B2C and B2B companies select the right partners and services to grow your ecommerce business to the next level.

We created Bullseye because a great many ecommerce businesses come to us for help in making their sites work for their organizations. They’ve hired agencies to build sites for them, but the end products are beyond their grasp in some ways, or not otherwise geared for the people and processes that exist within their organizations. Why have a sophisticated marketing plug-in if you don’t have a marketing department? Why select the most advanced technology available if you don’t have the necessary skills in house to run it?

It’s All About Right-Sizing

The goal of Bullseye is to assess the operational maturity of your business, which in turn dictates the complexity of solutions required to support sustained growth.

Maturity has nothing to do with your business, ecommerce or merchandising acumen. You can have that in spades and still be “immature” when it comes to selecting or optimizing your tech stack. Maturity has to do with your organization’s availability of time and resources to dedicate to an implementation, along with all existing systems touch your project.

A Quick & Easy Quiz

Bullseye is a quick and easy quiz that asks you some questions about your staff levels and skills, along with existing infrastructure components. This allows Something Digital to assess the products and partners that will work seamlessly with your people, processes and tech, and most importantly, will allow you to scale up in order to support your growth goals.

How are we able to distill such critical insight into a quick and easy quiz? We launched Something Digital twenty years ago, and through those decades we’ve assessed the maturity levels of B2B and B2C companies in every sector and size. We’ve worked with companies that have a few dozen employees as well as Fortune 500 retailers. Assessing client maturity is part and parcel to what we do, and we know what to look for.

And while we know, internally, the importance of assessing a client’s maturity level, we wanted Bullseye to help clients understand that importance as well. We believe it will enable all parties to understand why we recommend specific vendors, solutions and approaches.

A Roadmap for Strategy

We at Something Digital passionately believe that the outcome of your business strategy is a series of tactics that are designed to promote a defined end goal, whether that’s attracting more visitors to your website, generating more revenue, lowering the cost of customer acquisition, or boosting average order value. Because tactics enable strategy, it’s important that all ecommerce business managers have strategies, services and partners that match their level of maturity and help them grow.

Growth tactics differ between B2B and B2C, which is why we have a Bullseye for each. Interested in taking the test? Contact us.

Written by: Phillip Jackson, Ecommerce Evangelist

Something Digital - magento

MySQL Best Practices for Upgrading to Magento 2.3

If you’re a member of the Magento community, you’ve probably heard that Magento has declared end of life for 2.1.x support to be June 2019 (that’s now!). As Magento 2.1 approaches end of life, we at SD have started the processing of upgrading a number of our clients from 2.1 to 2.3, the latest major version of Magento 2.

Because of the number of core data changes that come along with the upgrade from 2.1 to 2.2 alone (i.e. the move from serialized strings to JSON), we ran into a few issues during the initial process of deploying a 2.3 instance to a staging environment. The most challenging issue came in the form of a MySQL error during the setup:upgrade process. If you’re unfamiliar with this particular Magento command, it’s used to perform changes to the database that the application code will require. Essentially, modules can be created to make changes to preexisting database tables, or even create new ones for their own use. Because of the conversion of serialized data to JSON in many of Magento’s core tables, including customer and sales data, this process can take a very, very long time to execute depending on the amount of data in your database.

Just to take a step back for a second, I was recently fortunate enough to have the opportunity to attend on online Oracle course for MySQL Performance Tuning. At the very core of understanding how to tune your MySQL instance’s performance lies the requirement that you must establish a consistent baseline of which to work off of. Along with setting this baseline, comes the responsibility of making sure that test environments have a similar amount of data and load on them to accurately test any updates. Clearly for us, it is of the utmost importance to have near identical replication of production data in a staging environment for us to test (and benchmark) the process of converting these loads of data for these upgrades. For a number of our clients, this means making sure that we have enough anonymized customer and sales data to give us an accurate representation of how long these deployments will take, or if we’ll run into issues in a production environment.

And run into issues we did! Aside from the usual errors one would expect to run into caused by corrupted serialized data, or incompatible 3rd party modules, we ran into a rather bizarre (see: misleading) MySQL error during that setup:upgrade process on one of our Community Edition clients with a large amount of sales data. See the error below:

Warning: Error while sending QUERY packet. PID=14322 in /vendor/magento/zendframework1/library/Zend/Db/Statement/Pdo.php on line 228

And that’s all folks! No other exceptions, errors, warnings, clues. Nada. Digging into the code a little further, Magento was attempting to query the database, but when the MySQL instance unexpectedly threw an error during execution of the query, the application code wasn’t handling it. So, all we were left with was this somewhat vague error message from MySQL.

If you’re unfamiliar with this error, a quick Google search will point you in the direction of increasing the max_allowed_packets variable on your MySQL instance. Which is what we did… again and again until we maxed out the variable at 1G. So, what gives?

As it turns out, this particular staging instance’s MySQL instance had been greatly modified. Many variables were not only changed from the default, but drastically different than what was configured on production. Which might make sense right? Production experiences heavier loads than staging, duh. But if we go back to proper performance tuning and the goals we mentioned above, we should have remembered to not just replicate the data in staging, but the MySQL instance itself.

One system variable in particular had been significantly decreased in staging, and that was the wait_timeout variable. As defined in the official MySQL Documentation this variable defines “the number of seconds the server waits for activity on a non-interactive connection before closing it.”. We increased this variable as necessary, re-ran the deployment, and now everything worked– the deployment, specifically setup:upgrade, was successful. So again, what gives?

We came to understand that one of the queries – the query referenced in the error above – was actually timing out because it was waiting on another very large query to finish. We were able to replicate this behavior in a local environment when reducing this wait_timeout variable, running two competing queries at the same time, and constructing one of these to be very large and time consuming.

There is no way that we would have run into this issue prior to a production deployment had we not made an attempt at replicating the live MySQL instance and its data in a test environment. Overall, I think our team learned a valuable lesson about both the size of upgrading from 2.1 to 2.3, and the importance of data replication in a test environment.

Written by: Jeremy Dennen, Back End Developer

Google Shopping

Google Shopping Tips & Tricks Part 2: Campaign Configuration

Now that we’ve covered feed management, let’s discuss campaign configurations. When formulating a strategy, it’s important to consider how much time you have to devote to the initial configuration and ongoing optimizations. While agencies might love to organize campaigns as specifically as possible, in-house marketers with smaller teams may prefer simplicity and automation. What works for someone else may not work for you, so always consider the level of effort required before making any big decisions.

Smart Shopping

Google’s Smart Shopping campaigns are for anyone who is in a bit of a time crunch, prefers to automate the heavy lifting, and wants visibility across all shopping placements (search, YouTube, display, and Gmail).

After creating a new campaign, select “sales” as your goal and make sure that your feed (which should be linked from Merchant Center) and country of origin are set. At the bottom of the screen, you will see two subtypes in the AdWords platform:

After you choose “Smart Shopping,” the remaining options are fairly minimal. All campaigns use a “maximize conversion value” bidding strategy, so it is important to set your budget realistically. If you’ve run standard shopping or display campaigns in the past, I suggest using those budgets as a guideline for Smart Shopping. When in doubt, start low and increase over time. Another available option in the “bid settings” section is to “set a target return on ad spend.” Enabling this option right away can stunt your campaign and prevent Google from spending your entire daily budget. I recommend leaving this option unchecked and revisiting it after your campaign has run for at least one month.

After you’ve completed these steps, that’s it! You could technically call it a day, start collecting data, and closely monitoring ROAS.

Product Groups

One other area where you can make optimizations for Smart Shopping is under the “product groups” tab. By default, you’ll notice an enabled group called “All products” that contains everything you’ve included in your approved product feed. If you need to exclude certain products, simply edit the group.

This is where feed custom labels become very useful. Setting these columns to title, category, and/or sku makes it easy to find and isolate specific products. If the products you need aren’t available, you may need to add a column to your feed.

Testing

If you’re already running a standard Shopping campaign without optimal results, consider testing Smart Shopping. Isolate a small group of products, exclude them from your traditional campaign, and include them in your Smart campaign. Monitor the results for at least one month and make a determination based on performance.

Sometimes it makes sense to include certain products in a Smart Shopping campaign while managing others via Traditional Shopping. I occasionally use Traditional Shopping for bestselling products and Smart Shopping for the rest of the catalog. Depending on a client’s needs and performance standards, it may make sense to adopt some combination of the two.

Traditional Shopping

Traditional shopping is more difficult to manage but gives a greater level of control over bidding and optimizations. Since it’s impossible for me to assess which strategy is best for you without reviewing historical data and revenue goals, I’m going to give a brief overview of some common configurations and when you might want to use them:

  • Single product ad groups (SPAGs) are useful for anyone who needs product-level bidding control and has serious time to devote to management. If you have thousands of skus, adopt another strategy because this one will easily suck up all of your time. Instead of separating product groups by category, price, or brand, SPAGs use product IDs. Each product ID becomes its own product group with individual bid settings.
  • Campaigns organized around shopping intent give you control over how much you bid for certain types of users. A user who is already familiar with your brand and searches for a specific product has a high intent to buy, whereas a user who searches for something generic isn’t as good of a fit. By using your campaign’s priority settings and negative keywords, you can ensure that you bid most competitively on users who are likely to purchase.
  • Setting up campaigns by price and profit margin are two other solid strategies. The higher the profit margin, the more I’m willing to spend to send potential buyers to my site. It doesn’t make sense to pay the amount per click for products with wildly different return (unless customer lifetime value proves otherwise).

 

Traditional Shopping puts you in a good position to maximize return, discover new users, and grow your business, but it only works if you have time to manage it. When in doubt, start simple, optimize over time, and if all else fails, test Smart Shopping (or give Something Digital a call 😉).

Written by: Lindsay Pugh, Senior Digital Strategist