diff --git a/README.md b/README.md
index d14cec1..f34ebf0 100644
--- a/README.md
+++ b/README.md
@@ -70,6 +70,15 @@ Finally, initialize the processor:
 // logger is assumed to be a zerolog instance.
 processor := nanoq.NewProcessor(nanoq.NewClient(db), logger)
 
+// The default retry policy uses an exponential backoff with jitter,
+// but callers can provide their own if necessary.
+processor.RetryPolicy(func (t nanoq.Task) {
+	// First retry in 5s, every next retry in 1h.
+	if t.Retries == 0 {
+		return 5 * time.Second
+	}
+	return 1 * time.Hour
+})
 processor.OnError(func(ctx context.Context, t nanoq.Task, err error) {
 	// Log each failed task. 
 	// Idea: Send to Sentry when t.Retries == t.MaxRetries.
diff --git a/nanoq.go b/nanoq.go
index 674af6c..6324d1b 100644
--- a/nanoq.go
+++ b/nanoq.go
@@ -180,8 +180,11 @@ func (c *Client) ClaimTask(ctx context.Context, tx *sqlx.Tx) (Task, error) {
 	return t, nil
 }
 
-// UpdateTask updates the given task.
-func (c *Client) UpdateTask(ctx context.Context, tx *sqlx.Tx, t Task) error {
+// RetryTask schedules a retry of the given task.
+func (c *Client) RetryTask(ctx context.Context, tx *sqlx.Tx, t Task, retryIn time.Duration) error {
+	t.Retries++
+	t.ScheduledAt = time.Now().UTC().Add(retryIn)
+
 	_, err := tx.NamedExecContext(ctx, `UPDATE tasks SET retries = :retries, scheduled_at = :scheduled_at WHERE id = :id`, t)
 	if err != nil {
 		return err
@@ -247,8 +250,20 @@ type (
 
 	// Middleware wrap a handler in order to run logic before/after it.
 	Middleware func(next Handler) Handler
+
+	// RetryPolicy determines the retry delay for a given task.
+	RetryPolicy func(t Task) time.Duration
 )
 
+// DefaultRetryPolicy uses an exponential base delay with jitter.
+// Approximate examples: 7s, 50s, 5min, 20min, 50min, 2h, 4h, 9h, 16h, 27h.
+func DefaultRetryPolicy(t Task) time.Duration {
+	exp := 5 + int(math.Pow(float64(t.Retries+1), float64(5)))
+	s := exp + rand.IntN(exp/2)
+
+	return time.Duration(s) * time.Second
+}
+
 // Processor represents the queue processor.
 type Processor struct {
 	client *Client
@@ -257,6 +272,7 @@ type Processor struct {
 	errorHandler ErrorHandler
 	handlers     map[string]Handler
 	middleware   []Middleware
+	retryPolicy  RetryPolicy
 
 	done atomic.Bool
 }
@@ -264,10 +280,11 @@ type Processor struct {
 // NewProcessor creates a new processor.
 func NewProcessor(client *Client, logger zerolog.Logger) *Processor {
 	return &Processor{
-		client:     client,
-		logger:     logger,
-		handlers:   make(map[string]Handler),
-		middleware: make([]Middleware, 0),
+		client:      client,
+		logger:      logger,
+		handlers:    make(map[string]Handler),
+		middleware:  make([]Middleware, 0),
+		retryPolicy: DefaultRetryPolicy,
 	}
 }
 
@@ -290,6 +307,11 @@ func (p *Processor) Handle(taskType string, h Handler, ms ...Middleware) {
 	p.handlers[taskType] = h
 }
 
+// RetryPolicy registers the given retry policy.
+func (p *Processor) RetryPolicy(rp RetryPolicy) {
+	p.retryPolicy = rp
+}
+
 // Run starts the processor and blocks until a shutdown signal (SIGINT/SIGTERM) is received.
 //
 // Once the shutdown signal is received, workers stop claiming new tasks.
@@ -369,10 +391,8 @@ func (p *Processor) process(ctx context.Context) error {
 				p.errorHandler(ctx, t, err)
 			}
 			if t.Retries < t.MaxRetries && !errors.Is(err, ErrSkipRetry) {
-				t.Retries = t.Retries + 1
-				t.ScheduledAt = getNextRetryTime(int(t.Retries))
-
-				if err := p.client.UpdateTask(ctx, tx, t); err != nil {
+				retryIn := p.retryPolicy(t)
+				if err := p.client.RetryTask(ctx, tx, t, retryIn); err != nil {
 					return fmt.Errorf("update task %v: %w", t.ID, err)
 				}
 
@@ -411,14 +431,3 @@ func callHandler(ctx context.Context, h Handler, t Task) (err error) {
 
 	return h(ctx, t)
 }
-
-// getNextRetryTime returns the time of the next retry.
-// Uses an exponential base delay with jitter.
-// Approximate examples: 7s, 50s, 5min, 20min, 50min, 2h, 4h, 9h, 16h, 27h.
-func getNextRetryTime(n int) time.Time {
-	exp := 5 + int(math.Pow(float64(n), float64(5)))
-	s := exp + rand.IntN(exp/2)
-	d := time.Duration(s) * time.Second
-
-	return time.Now().UTC().Add(d)
-}
diff --git a/nanoq_test.go b/nanoq_test.go
index 619038a..e4a21c5 100644
--- a/nanoq_test.go
+++ b/nanoq_test.go
@@ -152,6 +152,11 @@ func TestProcessor_Run(t *testing.T) {
 	defer db.Close()
 	client := nanoq.NewClient(sqlx.NewDb(db, "sqlmock"))
 	processor := nanoq.NewProcessor(client, zerolog.Nop())
+	retryPolicyCalled := 0
+	processor.RetryPolicy(func(t nanoq.Task) time.Duration {
+		retryPolicyCalled++
+		return 1 * time.Second
+	})
 	processor.Handle("my-type", func(ctx context.Context, task nanoq.Task) error {
 		// Fail the task once.
 		if task.Retries == 0 {
@@ -199,6 +204,10 @@ func TestProcessor_Run(t *testing.T) {
 	if errorHandlerCalled != 1 {
 		t.Errorf("erorr handler called %v times instead of %v", errorHandlerCalled, 1)
 	}
+
+	if retryPolicyCalled != 1 {
+		t.Errorf("retry policy called %v times instead of %v", retryPolicyCalled, 1)
+	}
 }
 
 func TestProcessor_Run_RetriesExhausted(t *testing.T) {