Common Ecommerce Email Marketing Strategy Mistakes Hurting Sales and Retention

Some links on The Justifiable are affiliate links, meaning we may earn a small commission at no extra cost to you. Read full disclaimer.

The most common ecommerce email marketing strategy mistakes hurting sales usually look harmless at first: sending to everyone, relying on discounts, skipping important automations, or judging success by open rates.

Over time, however, these habits can reduce engagement, weaken customer relationships, and leave valuable repeat purchases unrealized.

A strong email program should do more than generate occasional campaign revenue. It should guide subscribers from first interest through purchase, repeat purchase, and long-term loyalty.

This guide shows you where ecommerce email strategies typically break down, how to diagnose the problems, and how to build a more sustainable approach.

Mistake 1: Treating Email As a Campaign Channel Instead of a Customer System

Many ecommerce brands begin with newsletters and promotions because campaigns are easy to understand. The problem starts when campaigns remain the entire strategy rather than becoming one part of a broader customer lifecycle system.

Relying Too Heavily on One-Off Promotional Campaigns

A promotional calendar can generate sales, but it is inherently reactive. You decide to launch a sale, choose an audience, build an email, and press send. Someone who visits your store the following week may receive an entirely different experience even though their behavior shows similar purchase intent.

Automated lifecycle messaging solves a different problem. It responds to what an individual shopper does: subscribing, viewing products, beginning checkout, buying for the first time, or becoming inactive.

That distinction matters because ecommerce customers rarely move through the funnel according to your campaign calendar.

A practical strategy usually combines two systems:

• Campaigns: Product launches, seasonal promotions, educational content, inventory announcements, and timely offers.
• Automations: Welcome, browsing, cart recovery, post-purchase, cross-sell, replenishment, review, and win-back journeys.

If 90% of your attention goes toward weekly campaigns while your automated journeys have not been reviewed for six months, your priorities are probably reversed.

I recommend treating campaigns as temporary opportunities and automations as permanent infrastructure. Build the infrastructure first. Then campaigns can create additional demand without carrying the entire burden of revenue generation.

Optimizing for the First Order Instead of the Customer Relationship

Another common mistake is designing every email around one question: “How do we make this person buy today?”

That approach can generate short-term revenue while weakening retention. Someone who just made a first purchase, for example, may need reassurance, product education, usage advice, or help choosing a complementary item. Sending another generic discount immediately can train the customer to associate your emails only with price reductions.

Map your email strategy around customer stages instead:

1. Subscriber who has never purchased.
2. High-intent visitor.
3. First-time customer.
4. Recent customer awaiting or using the product.
5. Repeat customer.
6. High-value or loyal customer.
7. Customer showing signs of inactivity.

The objective changes at each stage. Before a first purchase, reducing uncertainty may matter most. After purchase, improving the experience and encouraging successful product use may be more important. Later, complementary products or replenishment can become relevant.

This lifecycle view also stops you from asking every customer to take the same action. Retention improves when your messaging recognizes what the customer has already done and provides a logical next step rather than restarting the sales conversation every week.

Diagnosing Problems at the Lifecycle Level

Looking only at total email revenue can hide serious weaknesses. Imagine that promotional campaigns perform well during major sales events while new subscribers rarely convert and first-time customers seldom buy again. Your dashboard may still show respectable email revenue even though the underlying system is leaking customers.

Audit performance by lifecycle stage.

Ask:

• Are new subscribers becoming first-time buyers?
• Are cart abandoners returning to checkout?
• Are first-time customers progressing toward another purchase?
• Are recent purchasers receiving inappropriate promotional messages?
• Are formerly active subscribers gradually disengaging?
• Are valuable repeat customers receiving distinctive treatment?

You do not necessarily need complex software to perform the first audit. Start with your ecommerce platform, email platform, and customer purchase data.

Once the store becomes more complex, a platform such as Omnisend can help centralize ecommerce campaigns, automated workflows, and behavioral segmentation. It is particularly relevant when you want email automation tied closely to store activity. A smaller store with limited traffic may not need every advanced capability immediately, however. The priority is fixing the lifecycle logic, not adding software for its own sake.

A healthy ecommerce email strategy does not simply send more messages. It creates the right next message for each stage of the customer relationship.

Mistake 2: Sending the Same Emails to Too Much of Your List

Once the lifecycle is clear, audience selection becomes critical. Broad broadcasting is easier operationally, but sending every campaign to nearly everyone usually creates irrelevant experiences and makes performance harder to interpret.

Segmenting by Meaningful Customer Behavior

Basic segmentation often stops at “customers” and “non-customers.” That is better than nothing, but ecommerce stores usually have much richer behavioral data available.

Useful segmentation signals include:

• Products or categories purchased.
• Number of previous orders.
• Recency of the last purchase.
• Browsing activity.
• Cart or checkout behavior.
• Email engagement.
• Average or typical order value.
• Geographic location where operationally relevant.
• Customer lifecycle stage.

The right segment depends on the decision you are trying to make.

Suppose you sell skincare. A customer who recently purchased a cleanser should not automatically receive the same product promotion as someone who has repeatedly browsed cleansers without buying. The existing customer may be a better candidate for complementary products or usage education.

Avoid creating dozens of segments merely because your software permits it. Each segment should change something meaningful: the offer, product selection, message, timing, or frequency.

A useful test is simple: if two segments will receive exactly the same email indefinitely, you probably do not need them separated yet. Segmentation earns its complexity only when you can act differently because of what the data tells you.

Ignoring Engagement When Deciding Who Receives Campaigns

List size can become a vanity metric. A database containing 100,000 contacts does not necessarily mean 100,000 people should receive your next promotion.

People lose interest. Some subscribers signed up for one discount months ago. Others have stopped shopping in the category entirely. Continually treating every old subscriber as an equally valuable campaign recipient makes the audience less responsive.

Create engagement tiers based on your actual sending frequency and purchase cycle. A store selling frequently replenished consumables will need different time windows from a furniture company where customers may naturally go months between purchases.

You might distinguish between:

• Recently engaged subscribers.
• Active customers.
• Less-engaged subscribers.
• Long-term inactive contacts.

Then adjust campaign frequency accordingly.

Do not use email opens as the sole engagement signal. Privacy features and technical factors can make opens less dependable as an isolated measure. Combine them with stronger actions such as clicks, purchases, site behavior, and recent customer activity where your systems support it.

The objective is not simply to shrink your audience. It is to stop assuming that historical permission automatically means unlimited relevance. Sending fewer but more appropriate campaigns can give you a clearer picture of genuine demand.

Failing to Use Exclusions and Suppression Logic

Segmentation determines who should receive something. Exclusions determine who should not.

This second half is frequently neglected.

Suppose someone purchases a product this morning. That evening, they receive a campaign offering 15% off the exact product they just bought. Nothing is technically broken, but the customer experience is poor. Someone else may receive a promotional campaign while midway through a carefully designed welcome sequence, producing multiple messages within hours.

Create suppression rules around circumstances where another message would be redundant or disruptive.

Depending on your store, consider excluding:

• Very recent purchasers from acquisition-oriented offers.
• Customers already inside specific high-frequency automations.
• People who recently received several messages.
• Customers who bought the promoted product when repeat purchase is unlikely.
• Long-term unengaged contacts from your highest-volume sends.
• Subscribers who do not have the appropriate marketing status for the message.

Platforms such as Klaviyo are useful when behavioral data, dynamic segments, and event-driven flows become central to the strategy. The trade-off is that sophisticated segmentation creates more opportunities for configuration mistakes. Document your exclusions and periodically test profiles against them instead of assuming complex logic is automatically correct.

Mistake 3: Building Incomplete or Poorly Timed Ecommerce Automations

Automation is powerful precisely because it runs continuously. That also makes mistakes expensive: a weak campaign disappears after one send, while a weak automation can deliver the wrong experience to thousands of future customers.

Using a Single Welcome Email Instead of a Welcome Journey

A welcome email has several jobs competing for limited attention. It may need to deliver an incentive, explain the brand, establish expectations, address objections, highlight products, and persuade someone to shop.

Trying to accomplish all of that in one message usually produces either an overloaded email or an incomplete introduction.

A short welcome sequence gives you room to progress logically. For example:

1. Deliver the promised signup incentive and establish what the brand offers.
2. Help the subscriber choose a product or category.
3. Address the main buying objection or uncertainty.
4. Introduce proof, differentiators, or useful education.
5. Present a final incentive reminder if an expiring offer genuinely applies.

Your exact sequence should reflect the buying decision. A low-cost impulse product may need a shorter journey than an expensive item requiring comparison and consideration.

Set exit or branching rules as well. Someone who purchases after email two should not necessarily continue receiving “make your first purchase” messaging.

The mistake is not having too few emails by some universal standard. It is failing to give a new subscriber enough relevant information to move from interest toward an informed purchase.

Treating Cart Abandonment as the Only High-Intent Automation

Abandoned-cart emails receive considerable attention because the shopper has already selected a product. But focusing solely on carts ignores earlier and later behaviors that can also support revenue.

Depending on your store and available tracking, useful journeys may include:

• Browse abandonment for meaningful product interest.
• Checkout recovery for shoppers further through the purchase process.
• Post-purchase education.
• Cross-sell based on previous purchases.
• Replenishment for products with a predictable consumption pattern.
• Win-back messages when repeat customers become inactive.

Do not build every automation simply because it appears in a template library. Each workflow should correspond to a real customer state and a realistic next action.

For instance, browse abandonment works poorly when every casual product-page visitor immediately receives aggressive purchase reminders. Consider intent strength, frequency limits, product category, and whether the visitor has already been contacted elsewhere.

Automation is most effective when triggers represent meaningful changes in customer behavior. The goal is not to react to every trackable event. It is to recognize the events that indicate a useful opportunity to help the shopper progress.

Forgetting That Timing Changes the Meaning of the Message

The same email can feel useful or irritating depending on when it arrives.

Imagine a customer orders a gift. Before the item has even shipped, your automation asks them to buy another. The offer may be relevant eventually, but the timing ignores the customer’s immediate priority: confidence that the first order will arrive successfully.

Think in terms of customer context rather than arbitrary delays.

Before choosing a delay, ask:

• What is likely happening with the customer at this point?
• Has enough time passed for them to use the product?
• Could fulfilment delays make this message premature?
• Is the purchase cycle days, weeks, or months?
• What other emails might arrive during the same period?

For post-purchase flows, operational data can improve timing considerably. A review request after confirmed delivery makes more sense than one triggered blindly a fixed number of days after checkout.

I suggest periodically experiencing your flows exactly as a customer would. Place a test order, abandon a cart, subscribe through your primary form, and document every message received. A diagram can look perfectly logical while the actual inbox experience reveals collisions, awkward delays, and redundant offers immediately.

Mistake 4: Training Customers to Expect Discounts and Generic Promotions

Good automation cannot compensate for weak messaging. If nearly every email says “sale ends tonight,” your segmentation may become more sophisticated while the customer’s reason to remain subscribed becomes increasingly narrow.

Making Discounts the Default Reason to Open Your Emails

Discounts are useful tools, but they become strategically dangerous when they replace positioning, education, and relevance.

Frequent discounts can influence customer behavior. Shoppers may delay purchases because they expect another promotion soon. Loyal customers may feel disadvantaged after paying full price. Your team may also lose visibility into whether customers want the product itself or simply respond when the price drops.

Create additional reasons to engage.

Depending on the product, emails can cover:

• How to choose between product options.
• How to use or maintain a purchase.
• New product applications.
• Restocks and launches.
• Complementary product combinations.
• Customer questions and buying objections.
• Relevant seasonal needs.
• Behind-the-product information that helps buyers evaluate quality.

This does not mean turning every ecommerce email into a long educational newsletter. Commercial intent is appropriate. The point is to make the offer one component of your marketing rather than your entire value proposition.

Reserve discounts for situations where the economics and customer context justify them. When every automated branch ends in another coupon, revisit the underlying persuasion strategy first.

Personalizing With Data That Does Not Improve Relevance

Adding a first name is technically personalization, but it rarely changes the substance of a message.

Useful ecommerce personalization should answer a more important question: what do you know about this person’s interests or customer relationship that should change what you show them?

Examples include:

• Showing categories related to browsing behavior.
• Recommending complementary products based on previous orders.
• Adjusting messaging for first-time and repeat buyers.
• Avoiding promotions for something already purchased.
• Changing replenishment timing based on the likely product lifecycle.

Be cautious with personalization when the underlying data is incomplete. Incorrect recommendations can feel stranger than a well-written generic email.

Start with high-confidence information. A completed order is usually a stronger signal than one brief page view. Multiple related product interactions are stronger than an accidental click.

Build from broad relevance toward sophisticated personalization instead of creating elaborate rules immediately. One strong distinction—such as existing customer versus prospect—can improve messaging more than twenty fragile conditions that nobody on the team remembers six months later.

Personalization should reduce the distance between what the customer currently needs and what the email offers. If it does not change that relationship, it is mostly decoration.

Writing Around Your Promotion Instead of the Customer’s Decision

Many ecommerce emails are planned backward. The business decides what it wants to sell, then asks marketing to create a message around it.

Sometimes that is necessary, especially for launches or inventory priorities. But customer-centered email begins with the decision the recipient is making.

For a hesitant first-time buyer, the barrier might be uncertainty about fit, quality, shipping, returns, or product selection. For an existing customer, the question may be whether another item complements what they already own.

Build emails around those decisions.

A simple framework is:

1. Identify the customer state.
2. Identify the next useful action.
3. Identify the main obstacle preventing that action.
4. Address the obstacle.
5. Present the product or offer as the logical next step.

Consider a hypothetical apparel store. Instead of sending “20% Off Jackets” to every subscriber, it could send product-comparison guidance to subscribers who have repeatedly browsed outerwear, while recent jacket customers receive complementary accessories or care information.

The promotional objective has not disappeared. It has been placed inside a more relevant decision process.

That is the difference between merely distributing offers and using email to help customers progress.

Mistake 5: Neglecting Deliverability, Consent, and Database Quality

Email performance begins before anyone reads the copy. Poor data practices, irrelevant sending, and weak subscriber management can undermine even excellent creative work.

Growing the List Without Caring Who Joins It

More subscribers are useful only when they represent real people who genuinely want relevant communication.

Aggressive signup tactics can produce impressive list-growth charts while attracting low-intent subscribers primarily interested in an immediate incentive. The problem becomes worse if expectations are unclear about what happens after signup.

Your acquisition process should explain the value of joining. That value may include offers, early product access, educational content, product updates, or another legitimate benefit.

Track subscriber quality by source. Compare what happens after people enter through different forms, campaigns, partnerships, or acquisition channels.

Look beyond raw signup volume. Ask whether subscribers from each source:

• Confirm or complete signup where your process requires it.
• Engage with subsequent emails.
• Visit the store.
• Make a first purchase.
• Remain active over time.
• Unsubscribe or complain disproportionately.

A form generating half as many subscribers may be more valuable if those subscribers consistently become customers.

Your list-building metric should therefore connect acquisition with downstream behavior. Otherwise, you can spend money making the database larger while making the average subscriber less valuable.

Keeping Inactive Contacts in Your Main Sending Audience Forever

Marketers understandably hesitate to reduce campaign reach. An inactive contact can feel like a future sale waiting to happen.

But inactivity contains information.

Someone who has ignored months of campaigns, stopped visiting the store, and made no recent purchases should not necessarily receive messages at the same frequency as an active subscriber.

Create a re-engagement process before dramatically reducing contact. Give inactive subscribers a clear reason to return and test whether they still show interest.

Your rules should account for the natural purchase cycle. A customer buying mattresses is not “lost” merely because they have not purchased another mattress after 90 days. Email engagement and category behavior can provide context when repeat purchases naturally happen slowly.

After the re-engagement stage, consider suppressing contacts who remain inactive according to the criteria appropriate for your business.

The goal is not aggressive list pruning for its own sake. It is maintaining a mailing audience that reflects current interest rather than historical accumulation. Healthy list management also makes campaign metrics easier to interpret because you are not constantly averaging active customers together with years of dormant records.

Overlooking the Technical Foundation of Email Sending

Creative teams often focus on subject lines and design while technical sending configuration receives attention only after a deliverability problem appears.

Your sending setup should be treated as infrastructure.

Use a business sending domain and correctly configure the authentication settings supported or required by your email provider and domain environment. Maintain consistent sender identity, monitor delivery problems, and make unsubscribing straightforward rather than attempting to trap uninterested subscribers.

Also separate marketing messages from critical transactional communication where your systems require different handling. Customers must reliably receive information such as order confirmations and shipping-related updates regardless of how promotional engagement changes.

Watch for warning signs rather than waiting for a dramatic collapse:

• Increasing bounce levels.
• Falling click activity across normally strong segments.
• Sudden differences between mailbox providers.
• Unusual complaint levels.
• Large volumes of inactive addresses accumulating.
• Major performance changes after importing contacts or changing sending practices.

No email platform can permanently compensate for poor subscriber acquisition or irrelevant messaging. Deliverability is partly technical, but customer behavior matters too. A sustainable program starts with people who actually requested the communication and continues by giving them reasons to value it.

Mistake 6: Measuring Email With the Wrong Metrics

Once campaigns and automations are functioning, measurement determines what gets improved. Focusing on whichever metric looks most flattering can lead your team toward changes that do not increase customer value.

Treating Opens as the Main Definition of Success

Open rate can still provide contextual information, but it should not be your primary definition of ecommerce email success.

The action you ultimately care about normally happens later.

Measure performance according to the job of the email. For example:

Avoid interpreting metrics independently.

A campaign can generate a strong click-through rate but weak purchases because the landing page, pricing, product availability, or offer creates friction. Another campaign could produce significant attributed revenue primarily because it was sent to customers who were already preparing to buy.

Your measurement framework should therefore connect email behavior with store behavior.

Judge whether each workflow accomplishes the business task it was built to perform rather than optimizing every message toward the same superficial benchmark.

Confusing Attributed Revenue With Incremental Revenue

Email platforms commonly connect purchases back to marketing activity using an attribution model. This is useful, but attributed revenue does not automatically mean the email caused every recorded sale.

Consider a loyal customer who regularly buys every month. If they receive a promotional email shortly before their usual purchase, your reporting system may credit email with that order. The email may have influenced the timing, product choice, or order value—or the customer may have purchased anyway.

Treat attribution as a decision tool rather than absolute proof of causation.

Compare multiple signals:

• Email-platform reporting.
• Ecommerce revenue trends.
• Returning-customer behavior.
• Customer cohorts.
• Campaign timing.
• Controlled tests where feasible.
• Broader channel interactions.

As your marketing mix becomes more complicated, Triple Whale can provide additional ecommerce attribution and customer-journey analysis across marketing touchpoints. That becomes more useful when you are coordinating email with significant paid-media activity and need a broader measurement view. For a small store running only a few channels, additional attribution software may add cost and complexity before it adds clarity.

The principle matters more than the platform: use attribution to investigate contribution, not to claim certainty it cannot provide.

Testing Changes That Are Too Small to Matter

A/B testing becomes another vanity activity when teams continuously test minor wording differences without connecting experiments to meaningful hypotheses.

“Subject line A versus subject line B” can be useful, but you have much larger strategic variables to test.

Examples include:

• Different offers.
• Different audience segments.
• Message frequency.
• Education-first versus discount-first sequences.
• One-email versus multi-email flows.
• Product-specific versus category-level recommendations.
• Different delays inside automations.
• Different post-purchase pathways.

Write down the hypothesis before launching the test.

For example: “Customers who purchased product A will respond better to a complementary-product recommendation than to a storewide promotion during the first post-purchase month.”

Then decide what outcome would support or weaken the hypothesis.

Avoid changing five variables simultaneously. If the audience, creative, incentive, timing, and landing page all change, you may generate a winner without learning why it won.

Prioritize tests by potential business impact. Improving a major welcome or cart workflow can affect thousands of future customers. Testing an obscure campaign button color may affect almost nobody. Your testing queue should reflect that difference.

Mistake 7: Scaling Email Complexity Before Building a Reliable Operating System

As revenue grows, email programs accumulate segments, flows, campaigns, integrations, templates, and reporting dashboards. Scaling successfully requires more than adding automation; it requires keeping the system understandable and controllable.

Choosing Software Before Defining the Strategy

Email platforms are frequently evaluated by feature count. The more mature approach is to begin with what your store actually needs to accomplish.

Create a requirements list based on your strategy:

• Ecommerce platform integration.
• Behavioral event tracking.
• Dynamic segmentation.
• Lifecycle automation.
• Product recommendations.
• SMS or other channels when genuinely required.
• Reporting depth.
• Testing capabilities.
• Integration with your analytics stack.
• Workflow complexity your team can realistically maintain.

Then evaluate tools against those requirements.

Omnisend can make sense when ecommerce-focused campaigns and automation need to remain relatively approachable within one system. Klaviyo is relevant when detailed behavioral data, segmentation, and sophisticated lifecycle workflows are central to your program. Neither choice automatically creates a good email strategy.

Platform cost is also more than the subscription price. Consider migration effort, implementation time, team training, integration maintenance, and the operational burden of increasingly complicated workflows.

Do not buy sophistication several stages before you can use it. A well-designed set of five automations in a suitable platform can outperform dozens of poorly maintained flows inside a more powerful system.

Your strategy should dictate the technology stack rather than allowing a software feature list to dictate your strategy.

Failing to QA Automations After They Launch

“Set it and forget it” is one of the most damaging descriptions ever applied to marketing automation.

Automations should run automatically, but they should not be managed automatically.

Products change. Discounts expire. Inventory disappears. Website URLs move. Branding is updated. Customer behavior changes. Integrations break. Segmentation rules become outdated.

Create a recurring workflow audit.

For every important automation, verify:

1. The trigger still works correctly.
2. Eligibility and exclusion logic remain appropriate.
3. Links lead to the intended pages.
4. Discounts and promotional terms remain valid.
5. Product blocks populate correctly.
6. Personalization has suitable fallbacks.
7. Timing still matches operations.
8. Customers exit or branch appropriately after purchasing.
9. Mobile presentation remains usable.
10. Reporting still captures the outcomes you need.

Audit high-volume and revenue-critical workflows more frequently than minor programs.

Also maintain test profiles representing different customer states. One might simulate a new subscriber, another a first-time purchaser, another a repeat customer.

Testing isolated emails is not enough. Test complete journeys because most automation problems occur between messages—where triggers, branches, exclusions, and timing interact.

Adding More Messages Instead of Improving the System

Growth creates pressure to “do more email.” That can mean more campaigns, more flows, more segments, and more channels.

Volume is not the same as maturity.

Before increasing frequency, understand the marginal value of your existing sends. Ask whether additional messages create new customer value or simply compete with communication already being delivered.

Look for overlap.

A single customer might simultaneously qualify for:

• A promotional campaign.
• A product-browse reminder.
• An abandoned-cart sequence.
• A post-purchase flow.
• A loyalty message.
• An SMS campaign.

Each individual message may make sense when viewed separately. Together, they can create an exhausting experience.

Build a communication hierarchy. High-intent service or transactional needs generally require different priority from routine marketing. Purchase-triggered journeys may need to suppress generic campaigns temporarily. Promotional frequency can vary according to engagement.

As the program grows, assign ownership for lifecycle stages and document important logic. Complexity becomes dangerous when nobody can explain why a customer received a particular combination of messages.

The best sign that your email strategy is scaling is not that you have more automations. It is that increasingly complex customer behavior still produces a coherent experience.

Fix the Highest-Impact Email Mistake First

The common ecommerce email marketing strategy mistakes hurting sales rarely require you to rebuild everything simultaneously. Start by finding the biggest break in the customer journey.

If new subscribers are not converting, audit your welcome experience and first-purchase path. If customers buy once but seldom return, concentrate on post-purchase education, cross-selling, replenishment, and win-back logic. If campaigns are losing engagement, examine audience selection, frequency, list quality, and message relevance before sending more often.

Then improve measurement so you can tell whether the change actually helped.

A mature ecommerce email program connects customer behavior, timing, segmentation, useful messaging, automation, and measurement into one system. Build those foundations before adding more flows or tools. Your next step should be to map every email a customer can currently receive, identify the most damaging gap or collision, and fix that point before expanding the program further.

Juxhin

I’m Juxhin, the voice behind The Justifiable.
I’ve spent 6+ years building blogs, managing affiliate campaigns, and testing the messy world of online business. Here, I cut the fluff and share the strategies that actually move the needle — so you can build income that’s sustainable, not speculative.

Related posts:

Ecommerce Email Marketing Mistakes to Avoid: Are You Making These Revenue-Killing Errors? Ecommerce Email Marketing Strategy to Increase Online Store Sales: A Practical Plan Ecommerce Email Marketing Strategy for Converting Organic Traffic Into Customers Why Is My Ecommerce Email Marketing Strategy Not Working? A Diagnostic Guide Why Is My Ecommerce Email Marketing ROI So Low Despite Strong Open Rates? Where Should Signup Forms Fit in an Ecommerce Email Marketing Strategy? A Placement Guide